* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-29 0:38 rwhron
0 siblings, 0 replies; 25+ messages in thread
From: rwhron @ 2003-10-29 0:38 UTC (permalink / raw)
To: piggin; +Cc: linux-kernel
On 28 Oct 2003 10:24:36 +1100, Nick Piggin wrote:
> There is an #if 0 part in the patch. Could you try setting it to #if 1
> please.
Changing the #if 0 to #if 1 appears to affect AIM7 fserver the most.
Patch tested is very similar to:
http://marc.theaimsgroup.com/?l=linux-kernel&m=106729967301172&w=2
2.6.0-test9-as as-fix patch
2.6.0-test9-as1 as-fix patch with #if 1
Either way, that version of as-fix patch looks good for both
AIM7 and tiobench.
AIM7 dbase workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test9-as 32 604.1 314.6 133.7
2.6.0-test9-as1 32 599.6 317.0 142.2
2.6.0-test5 32 590.2 322.0 131.9
2.6.0-test9 32 316.2 601.2 128.5
2.6.0-test9-as 64 783.3 485.3 256.3
2.6.0-test9-as1 64 776.8 489.4 257.6
2.6.0-test5 64 774.3 490.9 256.6
2.6.0-test9 64 460.9 824.8 246.2
2.6.0-test9-as 96 868.5 656.6 386.4
2.6.0-test9-as1 96 864.1 660.0 391.0
2.6.0-test5 96 859.0 663.9 389.5
2.6.0-test9 96 507.5 1123.5 368.1
2.6.0-test9-as 128 914.1 831.8 510.7
2.6.0-test9-as1 128 910.7 834.8 521.1
2.6.0-test5 128 910.3 835.2 528.6
2.6.0-test9 128 536.6 1416.9 487.0
2.6.0-test9-as 160 952.4 998.0 635.8
2.6.0-test9-as1 160 949.3 1001.2 652.2
2.6.0-test5 160 741.5 1281.7 644.8
2.6.0-test9 160 522.4 1819.2 619.9
2.6.0-test9-as1 192 981.4 1162.1 779.1
2.6.0-test9-as 192 976.6 1167.8 774.2
2.6.0-test5 192 840.0 1357.8 776.0
2.6.0-test9 192 513.5 2220.8 739.3
2.6.0-test9-as 224 1000.6 1329.7 917.0
2.6.0-test9-as1 224 994.9 1337.3 904.8
2.6.0-test5 224 986.9 1348.2 911.8
2.6.0-test9 224 503.9 2640.3 853.4
2.6.0-test9-as 256 1010.6 1504.7 1036.3
2.6.0-test9-as1 256 1009.0 1507.1 1030.0
2.6.0-test5 256 999.5 1521.3 1029.3
2.6.0-test9 256 530.8 2864.8 975.5
At low number of tasks, #if 0 does more work. As the
number of tasks goes up, having #if 1 does more work.
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test9-as 4 114.6 211.6 38.2
2.6.0-test5 4 79.7 304.0 38.3
2.6.0-test9-as1 4 79.3 305.8 38.4
2.6.0-test9 4 43.0 564.4 43.6
2.6.0-test9-as 8 175.0 277.0 65.9
2.6.0-test5 8 129.8 373.5 66.3
2.6.0-test9-as1 8 128.8 376.5 65.0
2.6.0-test9 8 61.4 789.1 68.5
2.6.0-test9-as 12 208.4 349.0 93.6
2.6.0-test5 12 164.1 443.0 93.1
2.6.0-test9-as1 12 158.4 459.1 91.6
2.6.0-test9 12 70.4 1033.1 97.6
2.6.0-test9-as 16 190.4 509.3 111.8
2.6.0-test9-as1 16 184.7 525.1 114.2
2.6.0-test5 16 173.4 559.0 116.5
2.6.0-test9 16 100.7 962.8 135.5
2.6.0-test9-as1 20 213.3 568.2 143.3
2.6.0-test9-as 20 190.3 636.8 142.8
2.6.0-test5 20 180.5 671.5 143.0
2.6.0-test9 20 113.0 1072.6 160.0
2.6.0-test9-as1 24 219.2 663.5 169.8
2.6.0-test9-as 24 193.6 751.2 167.3
2.6.0-test5 24 189.1 769.0 176.0
2.6.0-test9 24 116.8 1245.6 163.3
2.6.0-test9-as1 28 223.7 758.6 195.5
2.6.0-test9-as 28 205.7 825.0 200.6
2.6.0-test5 28 204.6 829.2 197.5
2.6.0-test9 28 132.7 1278.5 190.2
2.6.0-test9-as1 32 228.1 850.0 224.8
2.6.0-test5 32 213.7 907.4 228.9
2.6.0-test9-as 32 209.3 926.6 225.9
2.6.0-test9 32 130.4 1486.6 209.1
The regression seen after test5 is gone with the most
recent version of as-fix.
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test9-as1 64 1952.3 190.8 169.6
2.6.0-test9-as 64 1936.9 192.3 170.2
2.6.0-test5 64 1888.8 197.2 168.3
2.6.0-test9 64 1050.5 354.6 165.0
2.6.0-test9-as 128 2154.9 345.7 337.2
2.6.0-test9-as1 128 2146.3 347.1 338.1
2.6.0-test5 128 2094.9 355.6 338.1
2.6.0-test9 128 1316.6 565.8 330.6
2.6.0-test9-as 192 2267.5 492.8 509.8
2.6.0-test9-as1 192 2260.9 494.2 508.6
2.6.0-test5 192 2256.0 495.3 507.8
2.6.0-test9 192 1399.0 798.7 502.9
2.6.0-test9-as1 256 2417.6 616.3 678.9
2.6.0-test5 256 2389.0 623.7 683.8
2.6.0-test9-as 256 2389.0 623.7 679.8
2.6.0-test9 256 1472.1 1012.1 672.8
2.6.0-test5 320 2457.9 757.7 856.6
2.6.0-test9-as 320 2440.5 763.1 851.0
2.6.0-test9-as1 320 2420.7 769.4 855.9
2.6.0-test9 320 1590.3 1171.1 841.0
2.6.0-test9-as1 384 2498.2 894.6 1030.3
2.6.0-test9-as 384 2487.2 898.6 1038.3
2.6.0-test5 384 2480.4 901.0 1036.5
2.6.0-test9 384 1626.5 1374.1 1022.9
2.6.0-test5 448 2529.8 1030.7 1209.2
2.6.0-test9-as 448 2514.9 1036.8 1213.3
2.6.0-test9-as1 448 2499.6 1043.1 1206.1
2.6.0-test9 448 1636.7 1593.0 1199.5
2.6.0-test9-as 512 2598.7 1146.7 1390.1
2.6.0-test9-as1 512 2581.0 1154.5 1382.3
2.6.0-test5 512 2566.2 1161.2 1386.2
2.6.0-test9 512 1650.9 1805.0 1377.2
Tiobench below is not too exciting. The main thing is the regression
seen in test[678] is gone.
tiobench-0.3.3
Unit information
================
File size = 8192 megabytes
Blk Size = 4096 bytes
Rate = megabytes per second
CPU% = percentage of CPU used during the test
Latency = milliseconds
Lat% = percent of requests that took longer than X seconds
CPU Eff = Rate divided by CPU% - throughput per cpu load
Sequential Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
---------------------- --- --------------------------------------------
2.6.0-test9-as 1 28.83 13.66% 0.404 753.63 0.00000
2.6.0-test5 1 28.55 13.89% 0.408 754.59 0.00000
2.6.0-test9-as1 1 28.47 13.76% 0.409 885.44 0.00000
2.6.0-test5 8 34.01 17.09% 2.620 834.85 0.00000
2.6.0-test9-as1 8 33.90 16.82% 2.650 943.85 0.00000
2.6.0-test9-as 8 30.59 15.02% 2.963 843.18 0.00000
2.6.0-test9-as1 16 31.26 15.30% 5.703 1078.85 0.00000
2.6.0-test5 16 30.13 15.04% 5.980 1337.69 0.00000
2.6.0-test9-as 16 26.53 13.00% 6.758 1036.38 0.00000
2.6.0-test9-as1 32 29.63 14.81% 12.057 1459.39 0.00000
2.6.0-test5 32 29.46 14.62% 11.954 1560.02 0.00000
2.6.0-test9-as 32 24.84 12.43% 14.389 1842.53 0.00000
2.6.0-test5 64 25.24 13.57% 26.901 1743.05 0.00000
2.6.0-test9-as1 64 22.77 12.34% 29.750 1783.29 0.00000
2.6.0-test9-as 64 21.53 11.64% 31.063 1901.78 0.00000
2.6.0-test5 128 10.41 9.76% 129.662 1820.78 0.00000
2.6.0-test9-as1 128 10.27 9.41% 130.097 2172.39 0.00000
2.6.0-test9-as 128 10.05 9.13% 133.312 1753.29 0.00000
2.6.0-test5 256 9.13 13.30% 305.097 1446.43 0.00000
2.6.0-test9-as 256 8.62 12.50% 322.473 1698.87 0.00000
2.6.0-test9-as1 256 8.57 12.63% 325.050 1653.00 0.00000
Random Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
---------------------- --- --------------------------------------------
2.6.0-test9-as1 1 0.98 0.99% 12.001 109.76 0.00000
2.6.0-test5 1 0.95 0.94% 12.373 114.89 0.00000
2.6.0-test9-as 1 0.81 0.80% 14.484 117.63 0.00000
2.6.0-test5 8 4.11 4.58% 19.433 137.74 0.00000
2.6.0-test9-as1 8 4.03 4.39% 20.775 177.47 0.00000
2.6.0-test9-as 8 3.61 3.79% 23.024 161.12 0.00000
2.6.0-test9-as1 16 4.58 3.69% 36.975 396.44 0.00000
2.6.0-test5 16 4.05 3.90% 42.424 236.49 0.00000
2.6.0-test9-as 16 3.33 3.02% 51.015 293.01 0.00000
2.6.0-test5 32 4.41 3.58% 70.910 516.28 0.00000
2.6.0-test9-as1 32 4.21 3.95% 80.533 451.20 0.00000
2.6.0-test9-as 32 3.26 3.13% 102.403 697.51 0.00000
2.6.0-test5 64 4.47 4.00% 137.277 1077.53 0.00000
2.6.0-test9-as 64 4.40 3.88% 134.358 1229.07 0.00000
2.6.0-test9-as1 64 4.40 3.74% 138.895 1233.53 0.00000
2.6.0-test5 128 4.86 4.83% 243.704 2287.09 0.00000
2.6.0-test9-as1 128 4.68 4.47% 243.701 2172.91 0.00000
2.6.0-test9-as 128 4.51 4.51% 249.319 2547.23 0.00000
2.6.0-test5 256 5.51 7.00% 427.503 2511.79 0.00000
2.6.0-test9-as 256 5.23 6.79% 450.963 3573.61 0.00000
2.6.0-test9-as1 256 5.01 6.61% 467.361 3305.41 0.00000
Sequential Writes ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
---------------------- --- --------------------------------------------
2.6.0-test5 1 56.24 41.13% 0.170 3207.90 0.00000
2.6.0-test9-as1 1 55.71 39.50% 0.172 2225.67 0.00000
2.6.0-test9-as 1 55.49 39.25% 0.173 1806.75 0.00000
2.6.0-test5 8 31.86 31.70% 2.356 15355.31 0.05655
2.6.0-test9-as1 8 31.66 30.95% 2.364 14477.90 0.06008
2.6.0-test9-as 8 31.10 30.23% 2.453 14880.66 0.06271
2.6.0-test9-as1 16 32.63 31.80% 4.445 28724.28 0.08045
2.6.0-test9-as 16 32.10 31.67% 4.459 41263.17 0.07834
2.6.0-test5 16 32.07 32.22% 4.371 29214.57 0.07829
2.6.0-test9-as1 32 32.79 32.88% 8.351 70993.57 0.10638
2.6.0-test5 32 30.73 34.74% 8.843 59335.31 0.10657
2.6.0-test9-as 32 30.03 33.06% 8.852 61766.24 0.10576
2.6.0-test5 64 30.27 36.46% 16.798 119951.55 0.14372
2.6.0-test9-as1 64 28.70 34.79% 17.693 114731.30 0.15449
2.6.0-test9-as 64 27.10 32.73% 18.207 120362.26 0.15745
2.6.0-test5 128 29.98 38.86% 32.259 230569.79 0.20509
2.6.0-test9-as1 128 29.84 37.86% 31.890 224444.62 0.20604
2.6.0-test9-as 128 29.25 37.18% 32.428 210201.96 0.21028
2.6.0-test5 256 33.49 48.84% 53.571 391954.31 0.24166
2.6.0-test9-as 256 32.18 44.97% 56.356 453839.45 0.24067
2.6.0-test9-as1 256 31.23 43.69% 57.305 431916.55 0.23952
Random Writes ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
---------------------- --- --------------------------------------------
2.6.0-test9-as 1 3.08 2.79% 0.582 25.12 0.00000
2.6.0-test5 1 2.86 2.88% 1.117 50.58 0.00000
2.6.0-test9-as1 1 2.63 2.52% 1.481 380.77 0.00000
2.6.0-test5 8 3.85 5.10% 4.885 861.30 0.00000
2.6.0-test9-as1 8 3.69 3.69% 7.495 1084.38 0.00000
2.6.0-test9-as 8 3.34 3.49% 11.409 1369.42 0.00000
2.6.0-test5 16 3.70 3.93% 17.617 1419.36 0.00000
2.6.0-test9-as1 16 3.59 3.90% 18.985 1561.92 0.00000
2.6.0-test9-as 16 3.43 3.33% 16.893 1458.50 0.00000
2.6.0-test5 32 3.72 5.11% 14.666 1191.07 0.00000
2.6.0-test9-as1 32 3.67 4.34% 13.096 1346.98 0.00000
2.6.0-test9-as 32 3.40 4.22% 16.468 1422.35 0.00000
2.6.0-test5 64 3.69 5.68% 14.445 1116.77 0.00000
2.6.0-test9-as1 64 3.53 5.15% 18.633 1450.11 0.00000
2.6.0-test9-as 64 3.39 4.76% 19.373 1659.40 0.00000
2.6.0-test5 128 3.72 6.48% 14.037 1354.59 0.00000
2.6.0-test9-as 128 3.44 5.97% 12.805 1776.43 0.00000
2.6.0-test9-as1 128 3.41 5.81% 15.659 1974.75 0.00000
2.6.0-test5 256 3.57 6.59% 13.418 1912.18 0.00000
2.6.0-test9-as1 256 3.48 6.76% 5.699 1468.80 0.00000
2.6.0-test9-as 256 3.43 6.53% 8.807 1795.06 0.00000
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-26 10:38 rwhron
@ 2003-10-27 23:49 ` Dave Olien
0 siblings, 0 replies; 25+ messages in thread
From: Dave Olien @ 2003-10-27 23:49 UTC (permalink / raw)
To: rwhron, piggin; +Cc: linux-kernel, akpm
[-- Attachment #1: Type: text/plain, Size: 1418 bytes --]
On Sun, Oct 26, 2003 at 05:38:29AM -0500, rwhron@earthlink.net wrote:
> 2.6.0-test8-mm1 doesn't have the AIM7 database regression
> that 2.6.0-test8 has. The AIM7 fileserver and shared
> workloads regression between test5-mm2 and test8-mm1 is
> about 18%. Mainline regression was about 38% between test5
> and test8 on AIM7 fileserver and shared workloads.
>
Yes, I've noticed this as well. This has been the pattern since
test5. Every mainline release (test6, test7, test8, AND test9)
has a large regression in as-iosched vs deadline scheduler performance.
The regression is considerably smaller when the mm[1-4] patches are
applied. See the attachment for a collection of data points, collected
running REAIM, an updated version of AIM7.
In the mm[1-4] broken-out patches lists, there is an O_SYNC-SPEEDUP-2.patch
that has been in all of the mm[1-4] patches since test5. The comments at
the head of that patch says this patch reduces the blocking of writers on i_sem.
This patch probably allows more writer concurrency. This benefits
only O_SYNC writes. Reaim does O_SYNC writes. My guess this patch
is what reduces the as-iosched regression for mm[1-4] patched kernels.
I imagine the O_SYNC-SPEEDUP is unlikely to get into any more TEST kernels.
But maybe it gives a hint is to what's going on with as-iosched.
Letting in more concurrent writers seems to reduce as-iosched's peformance
regression.
[-- Attachment #2: reaim_as_vs_deadline --]
[-- Type: text/plain, Size: 1737 bytes --]
Some data points (marked with XXX) haven't been collected yet.
This data has been collected on the Scalable Test Platform (STP)
8-way machines at OSDL.
REAIM is an updated version of AIM7. Information on REAIM and
performance data collected on 2, 4, and 8 way machines can be found at
http://developer.osdl.org/cliffw/
Hardware details of the STP 8-way machine are at the bottom of this page.
More details on each test run can be viewed by construcing a URL from the
STP ID #. For example, the first entry in the table can be seen at URL
http://khack.osdl.org/279446
Maximum Jobs Per Minute
Kernel Version IO Scheduler STP ID # Peak Load Quick Convergence
2.6.0-test5 Deadline 279446 8950 8542
As-iosched 279444 8785 8489
mm1 Deadline 279560 8303 7998
As-iosched 279558 8401 7935
mm2 Deadline 279929 8309 8063
As-iosched 281948 8224 7741
mm3 Deadline 281859 8438 8064
As-iosched 281813 8207 7879
mm4 Deadline 281814 8222 7991
As-iosched 281815 8417 7963
2.6.0-test6 Deadline 281816 8302 7949 ****
As-iosched 281817 6934 6945
mm1 Deadline 281864 8375 7930
As-iosched 281863 8163 7923
mm2 Deadline 281948 8224 7741
As-iosched 281865 8309 7616
mm3 Deadline 281868 8280 8022
As-iosched 281867 8348 8017
mm4 Deadline 281870 8512 7868
As-iosched 281869 8307 7749
2.6.0-test7 Deadline 281231 8413 8002 ****
As-iosched 282095 6957 6790
mm1 Deadline 281548 8567 8033
As-iosched 282181 8346 7755
2.6.0-test8 Deadline 281671 8294 8020 ****
As-iosched 281669 7018 6926
mm1 Deadline 282183 8326 7997
As-iosched 282182 8347 XXX
2.6.0-test9 Deadline XXXX XXXX
As-iosched 282310 6868 6045 ****
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-27 23:33 John Hawkes
0 siblings, 0 replies; 25+ messages in thread
From: John Hawkes @ 2003-10-27 23:33 UTC (permalink / raw)
To: linux-kernel
I ran AIM7 on a 64p Altix NUMA (Itanium 2 Madison CPUs, 1.5 GHz), with an
XFS filesystem on each of 99 fiberchannel disks, with two different flavors
of AIM7 workloads, on both 2.6.0-test5 and 2.6.0-test8. One workload shows
that -test8 is half the throughput of -test5. The other workload shows
that -test8 is 22% faster.
The first workload is the default AIM7 "multiuser/shared system" workload.
2.6.0-test8 shows a peak throughput of 52% of -test5 -- it drops in half.
The -test8 peak at 64p is roughly the same as at 8p and 4p. This
bottleneck cannot be attributed solely to I/O hardware, since our SGI
ProPack kernel (based upon 2.4.21) achieves a peak throughput at 64p that
is 30% higher than 2.6.0-test5's peak at 64p.
The second workload takes the same "multiuser/shared system" workload,
but turns off the O_SYNC flag for the three sync_* subtests. This
generally produces higher peak throughput values than when using the
default O_SYNC, and the peak throughput values scale significantly better
as CPU counts increase. Here the -test8 kernel is 22% *faster* than
-test5 at 64p (and 8x faster for no-O_SYNC than with-O_SYNC).
John Hawkes
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-26 10:38 rwhron
2003-10-27 23:49 ` Dave Olien
0 siblings, 1 reply; 25+ messages in thread
From: rwhron @ 2003-10-26 10:38 UTC (permalink / raw)
To: linux-kernel; +Cc: akpm
2.6.0-test8-mm1 doesn't have the AIM7 database regression
that 2.6.0-test8 has. The AIM7 fileserver and shared
workloads regression between test5-mm2 and test8-mm1 is
about 18%. Mainline regression was about 38% between test5
and test8 on AIM7 fileserver and shared workloads.
AIM7 dbase workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5-mm2 32 609.3 311.9 139.5
2.6.0-test8-mm1 32 608.2 312.6 133.5
2.6.0-test5 32 590.2 322.0 131.9
2.6.0-test8 32 299.8 634.0 128.6
2.6.0-test5-mm2 64 786.2 483.5 253.2
2.6.0-test8-mm1 64 784.3 484.7 256.6
2.6.0-test5 64 774.3 490.9 256.6
2.6.0-test8 64 450.5 843.9 246.5
2.6.0-test8-mm1 96 876.2 650.8 384.0
2.6.0-test5-mm2 96 869.2 656.0 382.2
2.6.0-test5 96 859.0 663.9 389.5
2.6.0-test8 96 549.7 1037.4 366.0
2.6.0-test5-mm2 128 929.4 818.0 526.1
2.6.0-test8-mm1 128 924.7 822.2 523.3
2.6.0-test5 128 910.3 835.2 528.6
2.6.0-test8 128 558.7 1360.8 497.6
2.6.0-test5-mm2 160 962.5 987.4 674.6
2.6.0-test8-mm1 160 959.8 990.2 664.5
2.6.0-test5 160 741.5 1281.7 644.8
2.6.0-test8 160 522.1 1820.2 612.2
2.6.0-test5-mm2 192 993.9 1147.5 810.0
2.6.0-test8-mm1 192 983.7 1159.3 806.0
2.6.0-test5 192 840.0 1357.8 776.0
2.6.0-test8 192 515.5 2212.4 742.7
2.6.0-test5-mm2 224 1011.2 1315.8 943.5
2.6.0-test8-mm1 224 1011.1 1315.9 918.0
2.6.0-test5 224 986.9 1348.2 911.8
2.6.0-test8 224 529.9 2510.8 859.5
2.6.0-test5-mm2 256 1030.9 1475.1 1076.5
2.6.0-test8-mm1 256 1023.2 1486.2 1078.5
2.6.0-test5 256 999.5 1521.3 1029.3
2.6.0-test8 256 552.4 2753.0 980.4
-mm regressed about 20% going from test5 to test8 in AIM7
fileserver workload. Linus tree regressed about 38%.
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5 4 79.7 304.0 38.3
2.6.0-test5-mm2 4 77.4 313.4 38.3
2.6.0-test8-mm1 4 66.5 364.5 34.1
2.6.0-test8 4 42.3 573.6 40.9
2.6.0-test5 8 129.8 373.5 66.3
2.6.0-test5-mm2 8 103.1 470.4 62.0
2.6.0-test8-mm1 8 85.1 569.6 63.9
2.6.0-test8 8 63.3 765.7 71.6
2.6.0-test5 12 164.1 443.0 93.1
2.6.0-test5-mm2 12 112.4 646.8 98.4
2.6.0-test8-mm1 12 92.3 788.0 101.8
2.6.0-test8 12 83.2 874.2 95.0
2.6.0-test5 16 173.4 559.0 116.5
2.6.0-test5-mm2 16 128.8 752.9 119.9
2.6.0-test8-mm1 16 104.4 928.5 124.1
2.6.0-test8 16 87.5 1108.4 102.5
2.6.0-test5 20 180.5 671.5 143.0
2.6.0-test5-mm2 20 141.7 855.3 144.1
2.6.0-test8-mm1 20 116.6 1039.9 143.9
2.6.0-test8 20 105.0 1154.6 136.7
2.6.0-test5 24 189.1 769.0 176.0
2.6.0-test5-mm2 24 156.2 930.9 180.2
2.6.0-test8-mm1 24 127.1 1144.1 174.4
2.6.0-test8 24 111.2 1308.1 162.2
2.6.0-test5 28 204.6 829.2 197.5
2.6.0-test5-mm2 28 158.7 1069.2 209.8
2.6.0-test8-mm1 28 127.8 1327.5 203.8
2.6.0-test8 28 130.7 1297.9 185.8
2.6.0-test5 32 213.7 907.4 228.9
2.6.0-test5-mm2 32 151.9 1276.7 239.7
2.6.0-test8-mm1 32 123.2 1573.5 243.7
2.6.0-test8 32 134.2 1444.5 212.9
On the shared workload, Linus tree regressed about 38% between
test5 and test8. The -mm tree only regressed about 16%.
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5 64 1888.8 197.2 168.3
2.6.0-test5-mm2 64 1517.6 245.4 177.0
2.6.0-test8-mm1 64 1265.2 294.4 174.8
2.6.0-test8 64 829.1 449.3 167.8
2.6.0-test5 128 2094.9 355.6 338.1
2.6.0-test5-mm2 128 1837.4 405.4 352.4
2.6.0-test8-mm1 128 1558.3 478.1 348.9
2.6.0-test8 128 1105.2 674.1 330.5
2.6.0-test5 192 2256.0 495.3 507.8
2.6.0-test5-mm2 192 2074.9 538.5 530.9
2.6.0-test8-mm1 192 1755.2 636.6 526.9
2.6.0-test8 192 1320.0 846.6 504.7
2.6.0-test5 256 2389.0 623.7 683.8
2.6.0-test5-mm2 256 2207.2 675.0 720.6
2.6.0-test8-mm1 256 1855.2 803.1 708.4
2.6.0-test8 256 1462.1 1019.0 672.7
2.6.0-test5 320 2457.9 757.7 856.6
2.6.0-test5-mm2 320 2327.9 800.0 919.4
2.6.0-test8-mm1 320 1958.4 951.0 894.6
2.6.0-test8 320 1573.0 1184.0 845.2
2.6.0-test5 384 2480.4 901.0 1036.5
2.6.0-test5-mm2 384 2440.2 915.9 1111.6
2.6.0-test8-mm1 384 1997.9 1118.6 1087.2
2.6.0-test8 384 1601.1 1395.9 1028.7
2.6.0-test5 448 2529.8 1030.7 1209.2
2.6.0-test5-mm2 448 2510.0 1038.8 1306.1
2.6.0-test8-mm1 448 2030.0 1284.4 1279.6
2.6.0-test8 448 1659.7 1571.0 1205.2
2.6.0-test5 512 2566.2 1161.2 1386.2
2.6.0-test5-mm2 512 2579.1 1155.4 1504.0
2.6.0-test8-mm1 512 2046.4 1456.1 1481.3
2.6.0-test8 512 1614.4 1845.8 1381.5
The big increase in max latency on tiobench sequential reads at
256 threads affects both test8 and test8-mm1.
tiobench-0.3.3
File size = 8192 megabytes
Blk Size = 4096 bytes
Sequential Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------ --- --------------------------------------------
2.6.0-test5 1 28.55 13.89% 0.408 754.59 0.00000
2.6.0-test5-mm2 1 28.66 13.98% 0.406 742.76 0.00000
2.6.0-test8 1 28.66 13.57% 0.406 1010.39 0.00000
2.6.0-test8-mm1 1 28.53 13.79% 0.408 977.75 0.00000
2.6.0-test5 8 34.01 17.09% 2.620 834.85 0.00000
2.6.0-test5-mm2 8 32.44 16.44% 2.765 844.99 0.00000
2.6.0-test8 8 20.83 9.82% 4.437 1626.42 0.00000
2.6.0-test8-mm1 8 21.31 10.27% 4.334 1669.82 0.00000
2.6.0-test5 16 30.13 15.04% 5.980 1337.69 0.00000
2.6.0-test5-mm2 16 29.71 14.97% 6.024 1115.95 0.00000
2.6.0-test8 16 15.88 7.53% 11.604 2888.67 0.00000
2.6.0-test8-mm1 16 16.40 7.93% 11.203 2954.75 0.00000
2.6.0-test5 32 29.46 14.62% 11.954 1560.02 0.00000
2.6.0-test5-mm2 32 28.01 14.26% 12.808 1627.00 0.00000
2.6.0-test8 32 16.17 7.55% 22.170 5119.98 0.00000
2.6.0-test8-mm1 32 15.78 7.74% 23.080 4826.59 0.00000
2.6.0-test5 64 25.24 13.57% 26.901 1743.05 0.00000
2.6.0-test5-mm2 64 24.01 13.27% 27.922 1849.82 0.00000
2.6.0-test8 64 8.71 4.78% 82.514 11744.03 3.04523
2.6.0-test8-mm1 64 9.06 5.13% 79.839 9458.66 2.92964
2.6.0-test5 128 10.41 9.76% 129.662 1820.78 0.00000
2.6.0-test5-mm2 128 11.00 10.44% 120.118 1742.44 0.00000
2.6.0-test8 128 3.04 2.64% 466.928 26583.72 8.12159
2.6.0-test8-mm1 128 3.00 2.78% 476.711 22323.98 8.28352
2.6.0-test5 256 9.13 13.30% 305.097 1446.43 0.00000
2.6.0-test5-mm2 256 9.09 13.60% 304.019 1648.80 0.00000
2.6.0-test8 256 2.66 2.41% 960.563 679406.45 9.02886
2.6.0-test8-mm1 256 2.80 2.61% 916.388 678138.20 8.60138
test8 and test8-mm1 have an increase in requests taking more than 2
seconds on tiobench random reads at 256 threads. -mm1 regressed less
than mainline between test5 and test8.
Random Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------ --- --------------------------------------------
2.6.0-test5 1 0.95 0.94% 12.373 114.89 0.00000
2.6.0-test5-mm2 1 0.90 0.88% 13.061 113.19 0.00000
2.6.0-test8 1 0.95 0.85% 12.360 113.96 0.00000
2.6.0-test8-mm1 1 0.89 0.88% 13.222 107.35 0.00000
2.6.0-test5 8 4.11 4.58% 19.433 137.74 0.00000
2.6.0-test5-mm2 8 3.94 4.03% 20.931 144.16 0.00000
2.6.0-test8 8 3.52 3.59% 25.298 477.80 0.00000
2.6.0-test8-mm1 8 4.51 4.80% 19.777 299.43 0.00000
2.6.0-test5 16 4.05 3.90% 42.424 236.49 0.00000
2.6.0-test5-mm2 16 4.09 4.31% 41.651 255.17 0.00000
2.6.0-test8 16 3.92 3.56% 45.278 843.31 0.00000
2.6.0-test8-mm1 16 4.11 3.46% 42.043 608.63 0.00000
2.6.0-test5 32 4.41 3.58% 70.910 516.28 0.00000
2.6.0-test5-mm2 32 4.23 4.17% 78.492 372.63 0.00000
2.6.0-test8 32 4.25 3.45% 77.345 1324.85 0.00000
2.6.0-test8-mm1 32 3.70 3.69% 91.986 1270.38 0.00000
2.6.0-test5 64 4.47 4.00% 137.277 1077.53 0.00000
2.6.0-test5-mm2 64 4.79 4.53% 124.035 1144.38 0.00000
2.6.0-test8 64 3.57 3.32% 181.257 3029.47 0.00000
2.6.0-test8-mm1 64 3.66 3.24% 175.724 2493.85 0.00000
2.6.0-test5 128 4.86 4.83% 243.704 2287.09 0.00000
2.6.0-test5-mm2 128 5.08 4.85% 228.102 2391.21 0.00000
2.6.0-test8 128 3.28 3.08% 379.081 5007.00 0.00000
2.6.0-test8-mm1 128 3.40 3.52% 372.210 4671.71 0.00000
2.6.0-test5 256 5.51 7.00% 427.503 2511.79 0.00000
2.6.0-test5-mm2 256 5.23 7.39% 449.007 3169.28 0.00000
2.6.0-test8 256 2.07 2.40% 1162.555 13971.08 16.95313
2.6.0-test8-mm1 256 2.36 2.99% 1019.360 10935.31 10.65104
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-24 23:43 rwhron
0 siblings, 0 replies; 25+ messages in thread
From: rwhron @ 2003-10-24 23:43 UTC (permalink / raw)
To: piggin; +Cc: linux-kernel, akpm
Quad P3 Xeon. Only the as-iosched.c from test5
really fixes the regression.
2.6.0-test8 = vanilla
2.6.0-test8-as = vanilla + as-iosched.c from test5
2.6.0-test8-as2 = vanilla + first as-fix patch
2.6.0-test8-as = vanilla + second as-fix patch
AIM7 dbase workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 32 601.0 316.3 134.0
2.6.0-test8-as3 32 320.7 592.7 128.3
2.6.0-test8-as2 32 304.3 624.6 128.8
2.6.0-test8 32 299.8 634.0 128.6
2.6.0-test8-as 64 782.6 485.8 259.2
2.6.0-test8-as3 64 470.1 808.7 247.0
2.6.0-test8-as2 64 458.2 829.8 245.7
2.6.0-test8 64 450.5 843.9 246.5
2.6.0-test8-as 96 869.1 656.1 387.6
2.6.0-test8 96 549.7 1037.4 366.0
2.6.0-test8-as2 96 543.4 1049.4 370.0
2.6.0-test8-as3 96 521.4 1093.6 372.5
2.6.0-test8-as 128 919.3 827.0 522.8
2.6.0-test8-as2 128 564.1 1348.0 496.6
2.6.0-test8 128 558.7 1360.8 497.6
2.6.0-test8-as3 128 551.9 1377.6 499.9
2.6.0-test8-as 160 955.0 995.2 650.3
2.6.0-test8-as2 160 534.2 1779.2 618.2
2.6.0-test8 160 522.1 1820.2 612.2
2.6.0-test8-as3 160 517.0 1838.4 611.9
2.6.0-test8-as 192 981.9 1161.5 790.9
2.6.0-test8-as2 192 535.6 2129.3 746.5
2.6.0-test8-as3 192 530.1 2151.3 744.8
2.6.0-test8 192 515.5 2212.4 742.7
2.6.0-test8-as 224 1002.3 1327.5 910.2
2.6.0-test8-as3 224 531.7 2502.3 850.2
2.6.0-test8 224 529.9 2510.8 859.5
2.6.0-test8-as2 224 526.4 2527.5 852.9
2.6.0-test8-as 256 1009.9 1505.7 1027.7
2.6.0-test8-as3 256 555.3 2738.2 976.5
2.6.0-test8 256 552.4 2753.0 980.4
2.6.0-test8-as2 256 548.9 2770.4 975.3
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 4 80.5 301.0 37.0
2.6.0-test8-as2 4 46.4 521.9 42.5
2.6.0-test8-as3 4 43.8 553.0 44.5
2.6.0-test8 4 42.3 573.6 40.9
2.6.0-test8-as 8 133.0 364.6 64.0
2.6.0-test8-as3 8 65.5 740.3 71.1
2.6.0-test8 8 63.3 765.7 71.6
2.6.0-test8-as2 8 59.9 809.1 74.4
2.6.0-test8-as 12 166.7 436.3 93.4
2.6.0-test8 12 83.2 874.2 95.0
2.6.0-test8-as3 12 67.2 1081.6 99.8
2.6.0-test8-as2 12 62.6 1161.5 99.5
2.6.0-test8-as 16 175.9 551.2 111.5
2.6.0-test8 16 87.5 1108.4 102.5
2.6.0-test8-as2 16 84.1 1152.9 109.6
2.6.0-test8-as3 16 82.0 1182.4 112.0
2.6.0-test8-as 20 186.3 650.7 133.8
2.6.0-test8 20 105.0 1154.6 136.7
2.6.0-test8-as2 20 101.3 1196.4 135.8
2.6.0-test8-as3 20 100.5 1206.3 128.8
2.6.0-test8-as 24 202.8 717.3 161.2
2.6.0-test8-as2 24 121.4 1198.0 164.1
2.6.0-test8 24 111.2 1308.1 162.2
2.6.0-test8-as3 24 104.8 1387.9 163.9
2.6.0-test8-as 28 207.1 819.3 200.0
2.6.0-test8 28 130.7 1297.9 185.8
2.6.0-test8-as2 28 128.0 1326.1 190.3
2.6.0-test8-as3 28 123.4 1374.5 186.9
2.6.0-test8-as 32 210.6 920.7 226.5
2.6.0-test8 32 134.2 1444.5 212.9
2.6.0-test8-as3 32 130.4 1487.1 217.7
2.6.0-test8-as2 32 122.4 1583.7 217.5
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 64 1949.6 191.1 170.8
2.6.0-test8-as2 64 880.0 423.3 166.6
2.6.0-test8-as3 64 878.8 423.9 166.3
2.6.0-test8 64 829.1 449.3 167.8
2.6.0-test8-as 128 2148.3 346.8 339.2
2.6.0-test8-as3 128 1176.2 633.4 331.1
2.6.0-test8-as2 128 1128.7 660.0 331.4
2.6.0-test8 128 1105.2 674.1 330.5
2.6.0-test8-as 192 2265.7 493.2 507.1
2.6.0-test8-as3 192 1330.7 839.7 505.0
2.6.0-test8 192 1320.0 846.6 504.7
2.6.0-test8-as2 192 1317.8 848.0 503.3
2.6.0-test8-as 256 2382.9 625.3 678.0
2.6.0-test8-as3 256 1490.3 999.7 673.6
2.6.0-test8 256 1462.1 1019.0 672.7
2.6.0-test8-as2 256 1453.3 1025.2 670.9
2.6.0-test8-as 320 2450.5 760.0 853.7
2.6.0-test8-as3 320 1579.7 1178.9 849.0
2.6.0-test8 320 1573.0 1184.0 845.2
2.6.0-test8-as2 320 1564.0 1190.8 848.2
2.6.0-test8-as 384 2517.4 887.8 1030.6
2.6.0-test8-as3 384 1633.0 1368.6 1020.7
2.6.0-test8-as2 384 1609.8 1388.3 1026.2
2.6.0-test8 384 1601.1 1395.9 1028.7
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 0:37 rwhron
2003-10-20 1:35 ` Nick Piggin
2003-10-20 4:51 ` Dave Olien
@ 2003-10-24 13:52 ` Magnus Andersson
2 siblings, 0 replies; 25+ messages in thread
From: Magnus Andersson @ 2003-10-24 13:52 UTC (permalink / raw)
To: rwhron; +Cc: linux-kernel
On Sun, Oct 19, 2003 at 08:37:45PM -0400, rwhron@earthlink.net wrote:
> There was about a 50% regression in jobs/minute in AIM7
> database workload on quad P3 Xeon. The CPU time has not
> gone up, so the extra run time is coming from something
> else. (I/O or I/O scheduler?)
Hello all!
This sounds like the same problem I had with the 2.6 kernel.
I posted some stuff on the kernel-list about this.
Make a search for "2.6.0-test4 parallel seek & read problem"
on google to read my posts.
I think I ran into two different unrelated problems.
First problem was solved, but second problem is still there.
It is probably the same you are having.
My finding was that the problem is not the scheduler, but some
other part of the IO system.
1 seek & read issued by a program goes to the disk as 2 different
seeks & 2 reads, thus only giving 50 % performance.
I think the problem is that the elevators are flushed so often
so merges that should happend never happens.
/Magnus
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-24 12:46 rwhron
0 siblings, 0 replies; 25+ messages in thread
From: rwhron @ 2003-10-24 12:46 UTC (permalink / raw)
To: piggin; +Cc: linux-kernel
>> NOTPM was better with AS in test5? Does using as-iosched.c from test5
>> in a test8 kernel help?
> If that helps, can you try this patch.
That patch helps the AIM7 database load a little, but the
big regression is still there.
2.6.0-test8 = vanilla
2.6.0-test8-as = vanilla + as-iosched.c from test5
2.6.0-test8-asc2 = vanilla + patch.
AIM7 dbase workloa
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 32 601.0 316.3 134.0
2.6.0-test8-as2 32 304.3 624.6 128.8
2.6.0-test8 32 299.8 634.0 128.6
2.6.0-test8-as 64 782.6 485.8 259.2
2.6.0-test8-as2 64 458.2 829.8 245.7
2.6.0-test8 64 450.5 843.9 246.5
2.6.0-test8-as 96 869.1 656.1 387.6
2.6.0-test8 96 549.7 1037.4 366.0
2.6.0-test8-as2 96 543.4 1049.4 370.0
2.6.0-test8-as 128 919.3 827.0 522.8
2.6.0-test8-as2 128 564.1 1348.0 496.6
2.6.0-test8 128 558.7 1360.8 497.6
2.6.0-test8-as 160 955.0 995.2 650.3
2.6.0-test8-as2 160 534.2 1779.2 618.2
2.6.0-test8 160 522.1 1820.2 612.2
2.6.0-test8-as 192 981.9 1161.5 790.9
2.6.0-test8-as2 192 535.6 2129.3 746.5
2.6.0-test8 192 515.5 2212.4 742.7
2.6.0-test8-as 224 1002.3 1327.5 910.2
2.6.0-test8 224 529.9 2510.8 859.5
2.6.0-test8-as2 224 526.4 2527.5 852.9
2.6.0-test8-as 256 1009.9 1505.7 1027.7
2.6.0-test8 256 552.4 2753.0 980.4
2.6.0-test8-as2 256 548.9 2770.4 975.3
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 4 80.5 301.0 37.0
2.6.0-test8-as2 4 46.4 521.9 42.5
2.6.0-test8 4 42.3 573.6 40.9
2.6.0-test8-as 8 133.0 364.6 64.0
2.6.0-test8 8 63.3 765.7 71.6
2.6.0-test8-as2 8 59.9 809.1 74.4
2.6.0-test8-as 12 166.7 436.3 93.4
2.6.0-test8 12 83.2 874.2 95.0
2.6.0-test8-as2 12 62.6 1161.5 99.5
2.6.0-test8-as 16 175.9 551.2 111.5
2.6.0-test8 16 87.5 1108.4 102.5
2.6.0-test8-as2 16 84.1 1152.9 109.6
2.6.0-test8-as 20 186.3 650.7 133.8
2.6.0-test8 20 105.0 1154.6 136.7
2.6.0-test8-as2 20 101.3 1196.4 135.8
2.6.0-test8-as 24 202.8 717.3 161.2
2.6.0-test8-as2 24 121.4 1198.0 164.1
2.6.0-test8 24 111.2 1308.1 162.2
2.6.0-test8-as 28 207.1 819.3 200.0
2.6.0-test8 28 130.7 1297.9 185.8
2.6.0-test8-as2 28 128.0 1326.1 190.3
2.6.0-test8-as 32 210.6 920.7 226.5
2.6.0-test8 32 134.2 1444.5 212.9
2.6.0-test8-as2 32 122.4 1583.7 217.5
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 64 1949.6 191.1 170.8
2.6.0-test8-as2 64 880.0 423.3 166.6
2.6.0-test8 64 829.1 449.3 167.8
2.6.0-test8-as 128 2148.3 346.8 339.2
2.6.0-test8-as2 128 1128.7 660.0 331.4
2.6.0-test8 128 1105.2 674.1 330.5
2.6.0-test8-as 192 2265.7 493.2 507.1
2.6.0-test8 192 1320.0 846.6 504.7
2.6.0-test8-as2 192 1317.8 848.0 503.3
2.6.0-test8-as 256 2382.9 625.3 678.0
2.6.0-test8 256 1462.1 1019.0 672.7
2.6.0-test8-as2 256 1453.3 1025.2 670.9
2.6.0-test8-as 320 2450.5 760.0 853.7
2.6.0-test8 320 1573.0 1184.0 845.2
2.6.0-test8-as2 320 1564.0 1190.8 848.2
2.6.0-test8-as 384 2517.4 887.8 1030.6
2.6.0-test8-as2 384 1609.8 1388.3 1026.2
2.6.0-test8 384 1601.1 1395.9 1028.7
2.6.0-test8-as 448 2547.5 1023.5 1214.4
2.6.0-test8 448 1659.7 1571.0 1205.2
2.6.0-test8-as2 448 1656.4 1574.2 1201.6
2.6.0-test8-as 512 2594.6 1148.5 1389.7
2.6.0-test8-as2 512 1620.4 1839.0 1378.0
2.6.0-test8 512 1614.4 1845.8 1381.5
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-24 0:10 rwhron
0 siblings, 0 replies; 25+ messages in thread
From: rwhron @ 2003-10-24 0:10 UTC (permalink / raw)
To: piggin; +Cc: linux-kernel
> Randy can reproduce regressions, so I'll work on his first.
Below is test8 and test8 with the test5 as-iosched.c (2.6.0-test8-as).
Jobs/minute is way up, and total runtime is way down with the test5
as-iosched.c.
The machine (quad P3 xeon) locked up during a bonnie++ run. I can't
compare that to vanilla test8 because I stopped the test8 to run the
test5 as-iosched. Both of these were compiled with gcc-3.3.1 because
of the RedHat gcc-2.96 bug in test8. (In case the compiler was part
of the instability).
Earlier I ran 2.6.0-test3-mm2 compiled with gcc-3.3.1, and it
was stable on this box.
AIM7 dbase workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 32 601.0 316.3 134.0
2.6.0-test8 32 299.8 634.0 128.6
2.6.0-test8-as 64 782.6 485.8 259.2
2.6.0-test8 64 450.5 843.9 246.5
2.6.0-test8-as 96 869.1 656.1 387.6
2.6.0-test8 96 549.7 1037.4 366.0
2.6.0-test8-as 128 919.3 827.0 522.8
2.6.0-test8 128 558.7 1360.8 497.6
2.6.0-test8-as 160 955.0 995.2 650.3
2.6.0-test8 160 522.1 1820.2 612.2
2.6.0-test8-as 192 981.9 1161.5 790.9
2.6.0-test8 192 515.5 2212.4 742.7
2.6.0-test8-as 224 1002.3 1327.5 910.2
2.6.0-test8 224 529.9 2510.8 859.5
2.6.0-test8-as 256 1009.9 1505.7 1027.7
2.6.0-test8 256 552.4 2753.0 980.4
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 4 80.5 301.0 37.0
2.6.0-test8 4 42.3 573.6 40.9
2.6.0-test8-as 8 133.0 364.6 64.0
2.6.0-test8 8 63.3 765.7 71.6
2.6.0-test8-as 12 166.7 436.3 93.4
2.6.0-test8 12 83.2 874.2 95.0
2.6.0-test8-as 16 175.9 551.2 111.5
2.6.0-test8 16 87.5 1108.4 102.5
2.6.0-test8-as 20 186.3 650.7 133.8
2.6.0-test8 20 105.0 1154.6 136.7
2.6.0-test8-as 24 202.8 717.3 161.2
2.6.0-test8 24 111.2 1308.1 162.2
2.6.0-test8-as 28 207.1 819.3 200.0
2.6.0-test8 28 130.7 1297.9 185.8
2.6.0-test8-as 32 210.6 920.7 226.5
2.6.0-test8 32 134.2 1444.5 212.9
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test8-as 64 1949.6 191.1 170.8
2.6.0-test8 64 829.1 449.3 167.8
2.6.0-test8-as 128 2148.3 346.8 339.2
2.6.0-test8 128 1105.2 674.1 330.5
2.6.0-test8-as 192 2265.7 493.2 507.1
2.6.0-test8 192 1320.0 846.6 504.7
2.6.0-test8-as 256 2382.9 625.3 678.0
2.6.0-test8 256 1462.1 1019.0 672.7
2.6.0-test8-as 320 2450.5 760.0 853.7
2.6.0-test8 320 1573.0 1184.0 845.2
2.6.0-test8-as 384 2517.4 887.8 1030.6
2.6.0-test8 384 1601.1 1395.9 1028.7
2.6.0-test8-as 448 2547.5 1023.5 1214.4
2.6.0-test8 448 1659.7 1571.0 1205.2
2.6.0-test8-as 512 2594.6 1148.5 1389.7
2.6.0-test8 512 1614.4 1845.8 1381.5
I'm running Nick's patch
http://marc.theaimsgroup.com/?l=linux-kernel&m=106688130229337&w=2
on test8 at the moment. I should have some comparison numbers
within 24 hours.
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-23 20:35 ` Dave Olien
@ 2003-10-23 23:07 ` Nick Piggin
0 siblings, 0 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-23 23:07 UTC (permalink / raw)
To: Dave Olien; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
Dave Olien wrote:
>test5 didn't work with dbt2. The database run kept hanging.
>I've been running reaim, which doesn't have those problems with
>older kernels, but has shown similar performance problems.
>
>I need to also check again that reaim and dbt2 still show similar
>performance issues with newer kernel versions.
>
>I've lost my test machine for today. But I'll try test5 as-iosched
>in a test8 kernel with reaim tomorrow. If it works well, I'll try your
>following patch. Probably get it in the mail tomorrow afternoon.
>
Hi Dave,
Yeah thats right - I forgot. It was hanging because of as-iosched.c
too, so don't bother trying to run it :(
Anyway, Randy can reproduce regressions, so I'll work on his first.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-23 2:24 ` Nick Piggin
2003-10-23 3:48 ` Nick Piggin
@ 2003-10-23 20:35 ` Dave Olien
2003-10-23 23:07 ` Nick Piggin
1 sibling, 1 reply; 25+ messages in thread
From: Dave Olien @ 2003-10-23 20:35 UTC (permalink / raw)
To: Nick Piggin; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
test5 didn't work with dbt2. The database run kept hanging.
I've been running reaim, which doesn't have those problems with
older kernels, but has shown similar performance problems.
I need to also check again that reaim and dbt2 still show similar
performance issues with newer kernel versions.
I've lost my test machine for today. But I'll try test5 as-iosched
in a test8 kernel with reaim tomorrow. If it works well, I'll try your
following patch. Probably get it in the mail tomorrow afternoon.
Dave
On Thu, Oct 23, 2003 at 12:24:09PM +1000, Nick Piggin wrote:
>
>
> Dave Olien wrote:
>
> >Sorry, this patch didn't fix our performance problems. Mary just
> >finished running dbt2 on test8 with your patch:
> >
> >NOTPM kernel scheduler
> >965 2.6.0-test8-np AS
> >1632 2.6.-test6-mm4 deadline
> >
>
> Thanks. hmm.
> And NOTPM was better with AS in test5? Does using as-iosched.c from test5
> in a test8 kernel help?
>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-23 2:24 ` Nick Piggin
@ 2003-10-23 3:48 ` Nick Piggin
2003-10-23 20:35 ` Dave Olien
1 sibling, 0 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-23 3:48 UTC (permalink / raw)
To: Dave Olien; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
[-- Attachment #1: Type: text/plain, Size: 439 bytes --]
Nick Piggin wrote:
>
>
> Dave Olien wrote:
>
>> Sorry, this patch didn't fix our performance problems. Mary just
>> finished running dbt2 on test8 with your patch:
>>
>> NOTPM kernel scheduler
>> 965 2.6.0-test8-np AS
>> 1632 2.6.-test6-mm4 deadline
>>
>
> Thanks. hmm.
> And NOTPM was better with AS in test5? Does using as-iosched.c from test5
> in a test8 kernel help?
>
If that helps, can you try this patch.
[-- Attachment #2: as-fix.patch --]
[-- Type: text/plain, Size: 2416 bytes --]
linux-2.6-npiggin/drivers/block/as-iosched.c | 23 +++--------------------
1 files changed, 3 insertions(+), 20 deletions(-)
diff -puN drivers/block/as-iosched.c~as-fix drivers/block/as-iosched.c
--- linux-2.6/drivers/block/as-iosched.c~as-fix 2003-10-23 13:46:11.000000000 +1000
+++ linux-2.6-npiggin/drivers/block/as-iosched.c 2003-10-23 13:47:58.000000000 +1000
@@ -99,7 +99,6 @@ struct as_data {
sector_t last_sector[2]; /* last REQ_SYNC & REQ_ASYNC sectors */
struct list_head *dispatch; /* driver dispatch queue */
struct list_head *hash; /* request hash */
- unsigned long new_success; /* anticipation success on new proc */
unsigned long current_batch_expires;
unsigned long last_check_fifo[2];
int changed_batch; /* 1: waiting for old batch to end */
@@ -585,18 +584,11 @@ static void as_antic_stop(struct as_data
int status = ad->antic_status;
if (status == ANTIC_WAIT_REQ || status == ANTIC_WAIT_NEXT) {
- struct as_io_context *aic;
-
if (status == ANTIC_WAIT_NEXT)
del_timer(&ad->antic_timer);
ad->antic_status = ANTIC_FINISHED;
/* see as_work_handler */
kblockd_schedule_work(&ad->antic_work);
-
- aic = ad->io_context->aic;
- if (aic->seek_samples == 0)
- /* new process */
- ad->new_success = (ad->new_success * 3) / 4 + 256;
}
}
@@ -612,14 +604,8 @@ static void as_antic_timeout(unsigned lo
spin_lock_irqsave(q->queue_lock, flags);
if (ad->antic_status == ANTIC_WAIT_REQ
|| ad->antic_status == ANTIC_WAIT_NEXT) {
- struct as_io_context *aic;
ad->antic_status = ANTIC_FINISHED;
kblockd_schedule_work(&ad->antic_work);
-
- aic = ad->io_context->aic;
- if (aic->seek_samples == 0)
- /* new process */
- ad->new_success = (ad->new_success * 3) / 4;
}
spin_unlock_irqrestore(q->queue_lock, flags);
}
@@ -708,11 +694,10 @@ static int as_can_break_anticipation(str
return 1;
}
- if (ad->new_success < 256 &&
- (aic->seek_samples == 0 || aic->ttime_samples == 0)) {
+ if (aic->seek_samples == 0 || aic->ttime_samples == 0) {
/*
- * Process has just started IO and we have a bad history of
- * success anticipating on new processes!
+ * Process has just started IO. Don't anticipate.
+ * TODO! Must fix this up.
*/
return 1;
}
@@ -1822,8 +1807,6 @@ static int as_init(request_queue_t *q, e
if (ad->write_batch_count < 2)
ad->write_batch_count = 2;
- ad->new_success = 512;
-
return 0;
}
_
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-22 18:30 ` Dave Olien
2003-10-22 19:08 ` Dave Olien
@ 2003-10-23 2:24 ` Nick Piggin
2003-10-23 3:48 ` Nick Piggin
2003-10-23 20:35 ` Dave Olien
1 sibling, 2 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-23 2:24 UTC (permalink / raw)
To: Dave Olien; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
Dave Olien wrote:
>Sorry, this patch didn't fix our performance problems. Mary just
>finished running dbt2 on test8 with your patch:
>
>NOTPM kernel scheduler
>965 2.6.0-test8-np AS
>1632 2.6.-test6-mm4 deadline
>
Thanks. hmm.
And NOTPM was better with AS in test5? Does using as-iosched.c from test5
in a test8 kernel help?
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-22 18:30 ` Dave Olien
@ 2003-10-22 19:08 ` Dave Olien
2003-10-23 2:24 ` Nick Piggin
1 sibling, 0 replies; 25+ messages in thread
From: Dave Olien @ 2003-10-22 19:08 UTC (permalink / raw)
To: Nick Piggin; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
I don't know how useful this might but. I've been trying to
track down which patches to as-iosched resulted in the performance
drop. The results so far are a little confusing. I have been
running reaim on an 8-way system, comparing as-iosched performance
with deadline, on each kernel version. The numbers are "number
of jobs per minute". So, larger is better.
Deadline As-iosched
2.6.0-test5 8542 8589
2.6.0-test5-mm1 8303 8401
2.6.0-test5-mm2 8309 8224
2.6.0-test5-mm3 8222 8417
2.6.0-test6 8302 6934 ****
2.6.0-test6-mm1 8375 8163
2.6.0-test6-mm2 ??? 8309
I'm still getting data on test6-mm*. It's curious that
performance dropped for test6, but came back for test6-mm1.
On Wed, Oct 22, 2003 at 11:30:28AM -0700, Dave Olien wrote:
>
> Sorry, this patch didn't fix our performance problems. Mary just
> finished running dbt2 on test8 with your patch:
>
> NOTPM kernel scheduler
> 965 2.6.0-test8-np AS
> 1632 2.6.-test6-mm4 deadline
>
> This is an 8-way system with DAC960 and 12 LUNs, using raw devices.
> That's still quite a sizeable drop.
>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-22 9:54 ` Nick Piggin
@ 2003-10-22 18:30 ` Dave Olien
2003-10-22 19:08 ` Dave Olien
2003-10-23 2:24 ` Nick Piggin
0 siblings, 2 replies; 25+ messages in thread
From: Dave Olien @ 2003-10-22 18:30 UTC (permalink / raw)
To: Nick Piggin; +Cc: rwhron, venom, linux-kernel, akpm, Mary Edie Meredith
Sorry, this patch didn't fix our performance problems. Mary just
finished running dbt2 on test8 with your patch:
NOTPM kernel scheduler
965 2.6.0-test8-np AS
1632 2.6.-test6-mm4 deadline
This is an 8-way system with DAC960 and 12 LUNs, using raw devices.
That's still quite a sizeable drop.
On Wed, Oct 22, 2003 at 07:54:46PM +1000, Nick Piggin wrote:
>
> Could you please try the following patch against 2.6.0-test8.
> Thanks. I'll have to come up with something more comprehensive
> because this destroys find | xargs grep during other IO. It
> looks like this might be the cause of a lot of AS's database
> problems though.
>
> I'm working on something...
>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-21 13:05 rwhron
2003-10-21 15:57 ` Nick Piggin
2003-10-21 21:07 ` venom
@ 2003-10-22 9:54 ` Nick Piggin
2003-10-22 18:30 ` Dave Olien
2 siblings, 1 reply; 25+ messages in thread
From: Nick Piggin @ 2003-10-22 9:54 UTC (permalink / raw)
To: rwhron; +Cc: venom, linux-kernel, dmo, akpm
[-- Attachment #1: Type: text/plain, Size: 591 bytes --]
rwhron@earthlink.net wrote:
>>If you have time, would you please try testing as-iosched.c from
>>test5 in a later kernel (it won't go into test8-mm1 though).
>>
>
>copying drivers/block/as-iosched.c from 2.6.0-test5 to test8
>looks like it fixes the regression. Here are the results so far.
>
Could you please try the following patch against 2.6.0-test8.
Thanks. I'll have to come up with something more comprehensive
because this destroys find | xargs grep during other IO. It
looks like this might be the cause of a lot of AS's database
problems though.
I'm working on something...
[-- Attachment #2: as-fix.patch --]
[-- Type: text/plain, Size: 2416 bytes --]
linux-2.6-npiggin/drivers/block/as-iosched.c | 23 +++--------------------
1 files changed, 3 insertions(+), 20 deletions(-)
diff -puN drivers/block/as-iosched.c~as-fix drivers/block/as-iosched.c
--- linux-2.6/drivers/block/as-iosched.c~as-fix 2003-10-22 19:47:48.000000000 +1000
+++ linux-2.6-npiggin/drivers/block/as-iosched.c 2003-10-22 19:49:41.000000000 +1000
@@ -99,7 +99,6 @@ struct as_data {
sector_t last_sector[2]; /* last REQ_SYNC & REQ_ASYNC sectors */
struct list_head *dispatch; /* driver dispatch queue */
struct list_head *hash; /* request hash */
- unsigned long new_success; /* anticipation success on new proc */
unsigned long current_batch_expires;
unsigned long last_check_fifo[2];
int changed_batch; /* 1: waiting for old batch to end */
@@ -585,18 +584,11 @@ static void as_antic_stop(struct as_data
int status = ad->antic_status;
if (status == ANTIC_WAIT_REQ || status == ANTIC_WAIT_NEXT) {
- struct as_io_context *aic;
-
if (status == ANTIC_WAIT_NEXT)
del_timer(&ad->antic_timer);
ad->antic_status = ANTIC_FINISHED;
/* see as_work_handler */
kblockd_schedule_work(&ad->antic_work);
-
- aic = ad->io_context->aic;
- if (aic->seek_samples == 0)
- /* new process */
- ad->new_success = (ad->new_success * 3) / 4 + 256;
}
}
@@ -612,14 +604,8 @@ static void as_antic_timeout(unsigned lo
spin_lock_irqsave(q->queue_lock, flags);
if (ad->antic_status == ANTIC_WAIT_REQ
|| ad->antic_status == ANTIC_WAIT_NEXT) {
- struct as_io_context *aic;
ad->antic_status = ANTIC_FINISHED;
kblockd_schedule_work(&ad->antic_work);
-
- aic = ad->io_context->aic;
- if (aic->seek_samples == 0)
- /* new process */
- ad->new_success = (ad->new_success * 3) / 4;
}
spin_unlock_irqrestore(q->queue_lock, flags);
}
@@ -708,11 +694,10 @@ static int as_can_break_anticipation(str
return 1;
}
- if (ad->new_success < 256 &&
- (aic->seek_samples == 0 || aic->ttime_samples == 0)) {
+ if (aic->seek_samples == 0 && aic->ttime_samples == 0) {
/*
- * Process has just started IO and we have a bad history of
- * success anticipating on new processes!
+ * Process has just started IO. Don't anticipate.
+ * TODO! Must fix this up.
*/
return 1;
}
@@ -1822,8 +1807,6 @@ static int as_init(request_queue_t *q, e
if (ad->write_batch_count < 2)
ad->write_batch_count = 2;
- ad->new_success = 512;
-
return 0;
}
_
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-21 13:05 rwhron
2003-10-21 15:57 ` Nick Piggin
@ 2003-10-21 21:07 ` venom
2003-10-22 9:54 ` Nick Piggin
2 siblings, 0 replies; 25+ messages in thread
From: venom @ 2003-10-21 21:07 UTC (permalink / raw)
To: rwhron; +Cc: piggin, linux-kernel, dmo, akpm
sql bench on mysql, una master, three slave, with old as-iosched.c
runs mutch better, thanx.
On Tue, 21 Oct 2003 rwhron@earthlink.net wrote:
> Date: Tue, 21 Oct 2003 09:05:01 -0400
> From: rwhron@earthlink.net
> To: piggin@cyberone.com.au, venom@sns.it
> Cc: linux-kernel@vger.kernel.org, dmo@osdl.org, akpm@osdl.org
> Subject: Re: [BENCHMARK] I/O regression after 2.6.0-test5
>
> > If you have time, would you please try testing as-iosched.c from
> > test5 in a later kernel (it won't go into test8-mm1 though).
>
> copying drivers/block/as-iosched.c from 2.6.0-test5 to test8
> looks like it fixes the regression. Here are the results so far.
>
> In the AIM7 database benchmark, jobs/minute using test5 as-iosched.c
> doubled. At 32 tasks, jobs/min went from 300 to 601.
> At 256 tasks, jobs/min went from 552 to 1010.
>
> test5 as-iosched.c:
> cat suite7-dbase-2.6.0-test8-as
> Benchmark Version Machine Run Date
> AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
>
> Tasks Jobs/Min Real CPU
> 32 601.0 316.3 134.0
> 64 782.6 485.8 259.2
> 96 869.1 656.1 387.6
> 128 919.3 827.0 522.8
> 160 955.0 995.2 650.3
> 192 981.9 1161.5 790.9
> 224 1002.3 1327.5 910.2
> 256 1009.9 1505.7 1027.7
>
> Unmodififed test8.
> cat suite7-dbase-2.6.0-test8
> Benchmark Version Machine Run Date
> AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
>
> Tasks Jobs/Min Real CPU
> 32 299.8 634.0 128.6
> 64 450.5 843.9 246.5
> 96 549.7 1037.4 366.0
> 128 558.7 1360.8 497.6
> 160 522.1 1820.2 612.2
> 192 515.5 2212.4 742.7
> 224 529.9 2510.8 859.5
> 256 552.4 2753.0 980.4
>
> The AIM7 fileserver benchmark is back up to previous
> levels when test8 uses the test5 as-iosched.c:
>
>
> test5 as-iosched.c with test8.
> cat suite7-fserver-2.6.0-test8-as
> Benchmark Version Machine Run Date
> AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
>
> Tasks Jobs/Min Real CPU
> 4 80.5 301.0 37.0
> 8 133.0 364.6 64.0
> 12 166.7 436.3 93.4
> 16 175.9 551.2 111.5
> 20 186.3 650.7 133.8
> 24 202.8 717.3 161.2
> 28 207.1 819.3 200.0
> 32 210.6 920.7 226.5
>
>
> Unmodified test8.
> cat suite7-fserver-2.6.0-test8
> Benchmark Version Machine Run Date
> AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
>
> Tasks Jobs/Min Real CPU
> 4 42.3 573.6 40.9
> 8 63.3 765.7 71.6
> 12 83.2 874.2 95.0
> 16 87.5 1108.4 102.5
> 20 105.0 1154.6 136.7
> 24 111.2 1308.1 162.2
> 28 130.7 1297.9 185.8
> 32 134.2 1444.5 212.9
>
>
> My tiobench test hasn't completed yet, but test8 with test5's
> as-iosched.c is making good progress. I think Nick nailed the
> changeset.
>
> --
> Randy Hron
> http://home.earthlink.net/~rwhron/kernel/bigbox.html
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-21 13:05 rwhron
@ 2003-10-21 15:57 ` Nick Piggin
2003-10-21 21:07 ` venom
2003-10-22 9:54 ` Nick Piggin
2 siblings, 0 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-21 15:57 UTC (permalink / raw)
To: rwhron; +Cc: venom, linux-kernel, dmo, akpm
rwhron@earthlink.net wrote:
>>If you have time, would you please try testing as-iosched.c from
>>test5 in a later kernel (it won't go into test8-mm1 though).
>>
>
>copying drivers/block/as-iosched.c from 2.6.0-test5 to test8
>looks like it fixes the regression. Here are the results so far.
>
>In the AIM7 database benchmark, jobs/minute using test5 as-iosched.c
>doubled. At 32 tasks, jobs/min went from 300 to 601.
>At 256 tasks, jobs/min went from 552 to 1010.
>
Yep, looks right. Thanks for your thoroughness. There are a few
important direct IO bug fixes gone in there as well (hopefully
they aren't causing the regressions... unlikely), so for stability
you'll want to continue to use the test8 version.
I'll get a patch out soon.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-21 13:05 rwhron
2003-10-21 15:57 ` Nick Piggin
` (2 more replies)
0 siblings, 3 replies; 25+ messages in thread
From: rwhron @ 2003-10-21 13:05 UTC (permalink / raw)
To: piggin, venom; +Cc: linux-kernel, dmo, akpm
> If you have time, would you please try testing as-iosched.c from
> test5 in a later kernel (it won't go into test8-mm1 though).
copying drivers/block/as-iosched.c from 2.6.0-test5 to test8
looks like it fixes the regression. Here are the results so far.
In the AIM7 database benchmark, jobs/minute using test5 as-iosched.c
doubled. At 32 tasks, jobs/min went from 300 to 601.
At 256 tasks, jobs/min went from 552 to 1010.
test5 as-iosched.c:
cat suite7-dbase-2.6.0-test8-as
Benchmark Version Machine Run Date
AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
Tasks Jobs/Min Real CPU
32 601.0 316.3 134.0
64 782.6 485.8 259.2
96 869.1 656.1 387.6
128 919.3 827.0 522.8
160 955.0 995.2 650.3
192 981.9 1161.5 790.9
224 1002.3 1327.5 910.2
256 1009.9 1505.7 1027.7
Unmodififed test8.
cat suite7-dbase-2.6.0-test8
Benchmark Version Machine Run Date
AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
Tasks Jobs/Min Real CPU
32 299.8 634.0 128.6
64 450.5 843.9 246.5
96 549.7 1037.4 366.0
128 558.7 1360.8 497.6
160 522.1 1820.2 612.2
192 515.5 2212.4 742.7
224 529.9 2510.8 859.5
256 552.4 2753.0 980.4
The AIM7 fileserver benchmark is back up to previous
levels when test8 uses the test5 as-iosched.c:
test5 as-iosched.c with test8.
cat suite7-fserver-2.6.0-test8-as
Benchmark Version Machine Run Date
AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
Tasks Jobs/Min Real CPU
4 80.5 301.0 37.0
8 133.0 364.6 64.0
12 166.7 436.3 93.4
16 175.9 551.2 111.5
20 186.3 650.7 133.8
24 202.8 717.3 161.2
28 207.1 819.3 200.0
32 210.6 920.7 226.5
Unmodified test8.
cat suite7-fserver-2.6.0-test8
Benchmark Version Machine Run Date
AIM Multiuser Benchmark - Suite VII "1.1" dev4-003
Tasks Jobs/Min Real CPU
4 42.3 573.6 40.9
8 63.3 765.7 71.6
12 83.2 874.2 95.0
16 87.5 1108.4 102.5
20 105.0 1154.6 136.7
24 111.2 1308.1 162.2
28 130.7 1297.9 185.8
32 134.2 1444.5 212.9
My tiobench test hasn't completed yet, but test8 with test5's
as-iosched.c is making good progress. I think Nick nailed the
changeset.
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 7:55 ` venom
@ 2003-10-21 4:49 ` Nick Piggin
0 siblings, 0 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-21 4:49 UTC (permalink / raw)
To: venom; +Cc: Dave Olien, rwhron, linux-kernel, akpm
Hi Guys,
If you have time, would you please try testing as-iosched.c from
test5 in a later kernel (it won't go into test8-mm1 though).
Thanks
venom@sns.it wrote:
>me too, on some self made db benchs on both mysql and postgresql, testing, of
>course, I/O.
>
>On Sun, 19 Oct 2003, Dave Olien wrote:
>
>
>>Date: Sun, 19 Oct 2003 21:51:23 -0700
>>From: Dave Olien <dmo@osdl.org>
>>To: rwhron@earthlink.net
>>Cc: linux-kernel@vger.kernel.org, akpm@osdl.org
>>Subject: Re: [BENCHMARK] I/O regression after 2.6.0-test5
>>
>>
>>Yup, we've seen similar regression on tiobench and reaim workloads.
>>
>>On Sun, Oct 19, 2003 at 08:37:45PM -0400, rwhron@earthlink.net wrote:
>>
>>>There was about a 50% regression in jobs/minute in AIM7
>>>database workload on quad P3 Xeon. The CPU time has not
>>>gone up, so the extra run time is coming from something
>>>else. (I/O or I/O scheduler?)
>>>
>>>tiobench sequential reads has a significant regression too.
>>>
>>>Regression appears unrelated to filesystem type.
>>>
>>>dbench was not affected.
>>>
>>>The AIM7 was run on ext2.
>>>
>>>
>>>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 4:52 ` Andrew Morton
@ 2003-10-20 8:12 ` Nick Piggin
0 siblings, 0 replies; 25+ messages in thread
From: Nick Piggin @ 2003-10-20 8:12 UTC (permalink / raw)
To: Andrew Morton; +Cc: rwhron, linux-kernel
Andrew Morton wrote:
>Nick Piggin <piggin@cyberone.com.au> wrote:
>
>>
>>
>>rwhron@earthlink.net wrote:
>>
>>
>>>There was about a 50% regression in jobs/minute in AIM7
>>>database workload on quad P3 Xeon. The CPU time has not
>>>gone up, so the extra run time is coming from something
>>>else. (I/O or I/O scheduler?)
>>>
>>>tiobench sequential reads has a significant regression too.
>>>
>>>Regression appears unrelated to filesystem type.
>>>
>>>dbench was not affected.
>>>
>>>The AIM7 was run on ext2.
>>>
>>>
>>Yeah I'd say its all due to the IO scheduler. There is a problem
>>I'm thinking about how to fix - its the likely cause of this too.
>>
>>
>
>What change do you think it was due to?
>
>
I was thinking: [PATCH] fix AS crappy performance
(It still doesn't work properly)
>It's rather strange that test6 is slow but test6-mm is not: generally the
>IO scheduler regressions enter -mm first ;)
>
But if test6-mm isn't slow then maybe it is due to something else
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 4:51 ` Dave Olien
@ 2003-10-20 7:55 ` venom
2003-10-21 4:49 ` Nick Piggin
0 siblings, 1 reply; 25+ messages in thread
From: venom @ 2003-10-20 7:55 UTC (permalink / raw)
To: Dave Olien; +Cc: rwhron, linux-kernel, akpm
me too, on some self made db benchs on both mysql and postgresql, testing, of
course, I/O.
On Sun, 19 Oct 2003, Dave Olien wrote:
> Date: Sun, 19 Oct 2003 21:51:23 -0700
> From: Dave Olien <dmo@osdl.org>
> To: rwhron@earthlink.net
> Cc: linux-kernel@vger.kernel.org, akpm@osdl.org
> Subject: Re: [BENCHMARK] I/O regression after 2.6.0-test5
>
>
> Yup, we've seen similar regression on tiobench and reaim workloads.
>
> On Sun, Oct 19, 2003 at 08:37:45PM -0400, rwhron@earthlink.net wrote:
> > There was about a 50% regression in jobs/minute in AIM7
> > database workload on quad P3 Xeon. The CPU time has not
> > gone up, so the extra run time is coming from something
> > else. (I/O or I/O scheduler?)
> >
> > tiobench sequential reads has a significant regression too.
> >
> > Regression appears unrelated to filesystem type.
> >
> > dbench was not affected.
> >
> > The AIM7 was run on ext2.
> >
> > AIM7 dbase workload
> > kernel Tasks Jobs/Min Real CPU
> > 2.6.0-test5-mm2 32 609.3 311.9 139.5
> > 2.6.0-test5-mm4 32 595.2 319.4 131.7
> > 2.6.0-test6-mm4 32 590.8 321.8 130.7
> > 2.6.0-test5 32 590.2 322.0 131.9
> > 2.6.0-test6 32 320.6 592.9 129.1
> > 2.6.0-test7 32 290.3 654.8 120.9
> >
> > 2.6.0-test5-mm2 64 786.2 483.5 253.2
> > 2.6.0-test6-mm4 64 779.4 487.8 253.8
> > 2.6.0-test5 64 774.3 490.9 256.6
> > 2.6.0-test5-mm4 64 772.0 492.4 253.7
> > 2.6.0-test6 64 430.5 883.1 247.6
> > 2.6.0-test7 64 402.6 944.2 231.5
> >
> > 2.6.0-test5-mm2 96 869.2 656.0 382.2
> > 2.6.0-test6-mm4 96 867.6 657.2 387.4
> > 2.6.0-test5-mm4 96 863.1 660.7 385.2
> > 2.6.0-test5 96 859.0 663.9 389.5
> > 2.6.0-test6 96 478.5 1191.7 368.1
> > 2.6.0-test7 96 453.7 1256.9 346.8
> >
> > 2.6.0-test5-mm2 128 929.4 818.0 526.1
> > 2.6.0-test6-mm4 128 922.6 824.1 523.3
> > 2.6.0-test5-mm4 128 916.6 829.5 509.1
> > 2.6.0-test5 128 910.3 835.2 528.6
> > 2.6.0-test7 128 500.5 1519.1 461.3
> > 2.6.0-test6 128 500.2 1520.1 497.4
> >
> > 2.6.0-test5-mm2 160 962.5 987.4 674.6
> > 2.6.0-test5-mm4 160 952.7 997.6 655.5
> > 2.6.0-test6-mm4 160 950.0 1000.5 659.5
> > 2.6.0-test5 160 741.5 1281.7 644.8
> > 2.6.0-test7 160 518.5 1833.1 589.6
> > 2.6.0-test6 160 479.6 1981.8 618.9
> >
> > 2.6.0-test5-mm2 192 993.9 1147.5 810.0
> > 2.6.0-test5-mm4 192 982.7 1160.5 788.2
> > 2.6.0-test6-mm4 192 978.8 1165.2 808.4
> > 2.6.0-test5 192 840.0 1357.8 776.0
> > 2.6.0-test7 192 537.9 2120.2 709.3
> > 2.6.0-test6 192 494.3 2307.1 748.2
> >
> > 2.6.0-test5-mm2 224 1011.2 1315.8 943.5
> > 2.6.0-test5-mm4 224 1004.7 1324.3 942.1
> > 2.6.0-test6-mm4 224 999.5 1331.2 920.0
> > 2.6.0-test5 224 986.9 1348.2 911.8
> > 2.6.0-test7 224 546.7 2433.8 832.8
> > 2.6.0-test6 224 482.9 2755.1 862.1
> >
> > 2.6.0-test5-mm2 256 1030.9 1475.1 1076.5
> > 2.6.0-test5-mm4 256 1027.1 1480.6 1071.4
> > 2.6.0-test6-mm4 256 1010.3 1505.2 1074.8
> > 2.6.0-test5 256 999.5 1521.3 1029.3
> > 2.6.0-test7 256 556.6 2731.9 951.0
> > 2.6.0-test6 256 545.1 2789.5 978.4
> >
> >
> > Similarly, AIM7 fileserver benchmark had a drop in throughput.
> > CPU time is down, but total runtime is way up.
> >
> > AIM7 fserver workload
> > kernel Tasks Jobs/Min Real CPU
> > 2.6.0-test5 4 79.7 304.0 38.3
> > 2.6.0-test5-mm2 4 77.4 313.4 38.3
> > 2.6.0-test5-mm4 4 76.5 317.1 36.7
> > 2.6.0-test6-mm4 4 65.5 369.9 34.5
> > 2.6.0-test7 4 45.5 532.5 32.2
> > 2.6.0-test6 4 45.0 538.7 31.0
> >
> > 2.6.0-test5 8 129.8 373.5 66.3
> > 2.6.0-test5-mm2 8 103.1 470.4 62.0
> > 2.6.0-test5-mm4 8 102.2 474.2 61.8
> > 2.6.0-test6 8 82.7 585.9 84.5
> > 2.6.0-test6-mm4 8 82.4 588.1 61.1
> > 2.6.0-test7 8 61.0 795.0 59.9
> >
> > 2.6.0-test5 12 164.1 443.0 93.1
> > 2.6.0-test5-mm2 12 112.4 646.8 98.4
> > 2.6.0-test5-mm4 12 111.2 653.8 96.7
> > 2.6.0-test6-mm4 12 95.2 763.6 96.9
> > 2.6.0-test6 12 87.9 827.6 118.9
> > 2.6.0-test7 12 54.5 1335.0 91.8
> >
> > 2.6.0-test5 16 173.4 559.0 116.5
> > 2.6.0-test5-mm2 16 128.8 752.9 119.9
> > 2.6.0-test5-mm4 16 125.4 773.1 124.3
> > 2.6.0-test6-mm4 16 104.2 930.3 118.7
> > 2.6.0-test6 16 101.4 956.0 107.4
> > 2.6.0-test7 16 70.6 1372.6 112.2
> >
> > 2.6.0-test5 20 180.5 671.5 143.0
> > 2.6.0-test5-mm4 20 144.9 836.4 143.0
> > 2.6.0-test5-mm2 20 141.7 855.3 144.1
> > 2.6.0-test6 20 117.7 1030.0 131.1
> > 2.6.0-test6-mm4 20 114.4 1059.5 143.0
> > 2.6.0-test7 20 78.7 1539.9 134.0
> >
> > 2.6.0-test5 24 189.1 769.0 176.0
> > 2.6.0-test5-mm2 24 156.2 930.9 180.2
> > 2.6.0-test5-mm4 24 154.5 941.6 176.9
> > 2.6.0-test6 24 134.5 1081.1 156.5
> > 2.6.0-test6-mm4 24 124.5 1167.8 170.2
> > 2.6.0-test7 24 76.9 1892.1 169.0
> >
> > 2.6.0-test5 28 204.6 829.2 197.5
> > 2.6.0-test5-mm2 28 158.7 1069.2 209.8
> > 2.6.0-test5-mm4 28 156.0 1087.7 211.5
> > 2.6.0-test6-mm4 28 134.2 1264.0 196.9
> > 2.6.0-test6 28 133.7 1268.7 174.7
> > 2.6.0-test7 28 84.1 2016.4 188.5
> >
> > 2.6.0-test5 32 213.7 907.4 228.9
> > 2.6.0-test5-mm2 32 151.9 1276.7 239.7
> > 2.6.0-test5-mm4 32 151.5 1279.6 241.4
> > 2.6.0-test6 32 147.0 1318.9 197.0
> > 2.6.0-test6-mm4 32 138.8 1397.6 225.4
> > 2.6.0-test7 32 88.5 2190.3 214.9
> >
> > Same story for the shared workload. Jobs/minute is down a
> > lot since 2.6.0-test5.
> >
> >
> > AIM7 shared workload
> > kernel Tasks Jobs/Min Real CPU
> > 2.6.0-test5 64 1888.8 197.2 168.3
> > 2.6.0-test5-mm2 64 1517.6 245.4 177.0
> > 2.6.0-test5-mm4 64 1488.6 250.2 180.5
> > 2.6.0-test6-mm4 64 1239.0 300.6 175.0
> > 2.6.0-test7 64 1225.1 304.0 166.6
> > 2.6.0-test6 64 880.0 423.3 168.1
> >
> > 2.6.0-test5 128 2094.9 355.6 338.1
> > 2.6.0-test5-mm2 128 1837.4 405.4 352.4
> > 2.6.0-test5-mm4 128 1826.6 407.8 358.4
> > 2.6.0-test6-mm4 128 1561.8 477.0 346.6
> > 2.6.0-test7 128 1176.5 633.2 332.6
> > 2.6.0-test6 128 1142.1 652.3 332.6
> >
> > 2.6.0-test5 192 2256.0 495.3 507.8
> > 2.6.0-test5-mm2 192 2074.9 538.5 530.9
> > 2.6.0-test5-mm4 192 2044.8 546.5 536.9
> > 2.6.0-test6-mm4 192 1711.8 652.8 527.3
> > 2.6.0-test6 192 1328.3 841.3 505.6
> > 2.6.0-test7 192 1317.9 847.9 499.1
> >
> > 2.6.0-test5 256 2389.0 623.7 683.8
> > 2.6.0-test5-mm2 256 2207.2 675.0 720.6
> > 2.6.0-test5-mm4 256 2186.2 681.5 729.9
> > 2.6.0-test6-mm4 256 1780.7 836.7 719.0
> > 2.6.0-test6 256 1479.5 1007.0 672.9
> > 2.6.0-test7 256 1282.0 1162.2 672.8
> >
> > 2.6.0-test5 320 2457.9 757.7 856.6
> > 2.6.0-test5-mm4 320 2330.2 799.2 920.0
> > 2.6.0-test5-mm2 320 2327.9 800.0 919.4
> > 2.6.0-test6-mm4 320 1864.7 998.8 900.9
> > 2.6.0-test6 320 1575.9 1181.8 853.5
> > 2.6.0-test7 320 1317.9 1413.1 844.4
> >
> > 2.6.0-test5 384 2480.4 901.0 1036.5
> > 2.6.0-test5-mm2 384 2440.2 915.9 1111.6
> > 2.6.0-test5-mm4 384 2434.2 918.1 1111.0
> > 2.6.0-test6-mm4 384 1883.9 1186.3 1091.7
> > 2.6.0-test6 384 1629.8 1371.3 1026.3
> > 2.6.0-test7 384 1326.3 1685.0 1018.6
> >
> > 2.6.0-test5 448 2529.8 1030.7 1209.2
> > 2.6.0-test5-mm4 448 2511.6 1038.1 1306.8
> > 2.6.0-test5-mm2 448 2510.0 1038.8 1306.1
> > 2.6.0-test6-mm4 448 1934.3 1347.9 1283.6
> > 2.6.0-test6 448 1612.3 1617.1 1203.3
> > 2.6.0-test7 448 1344.7 1939.0 1200.9
> >
> > 2.6.0-test5-mm2 512 2579.1 1155.4 1504.0
> > 2.6.0-test5 512 2566.2 1161.2 1386.2
> > 2.6.0-test5-mm4 512 2552.9 1167.2 1509.2
> > 2.6.0-test6-mm4 512 1921.5 1550.8 1475.1
> > 2.6.0-test6 512 1538.6 1936.7 1376.9
> > 2.6.0-test7 512 1370.5 2174.3 1379.2
> >
> > tiobench read performance in test[67] has a big regression in throughput
> > as the number of threads goes up. The max and long latency numbers get
> > bad too. It appears to affect ext2, ext3, reiserfs, jfs and xfs.
> > Regression is in 2.6.0-test[67] and test6-mm4.
> >
> > tiobench-0.3.3
> > Unit information
> > ================
> > File size = 8192 megabytes
> > Blk Size = 4096 bytes
> > Rate = megabytes per second
> > CPU% = percentage of CPU used during the test
> > Latency = milliseconds
> > Lat% = percent of requests that took longer than X seconds
> > CPU Eff = Rate divided by CPU% - throughput per cpu load
> >
> > Sequential Reads ext2
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 28.55 13.89% 0.408 754.59 0.00000
> > 2.6.0-test5-mm2 1 28.66 13.98% 0.406 742.76 0.00000
> > 2.6.0-test5-mm4 1 28.46 13.90% 0.409 887.02 0.00000
> > 2.6.0-test6 1 28.47 13.81% 0.409 1069.74 0.00000
> > 2.6.0-test6-mm4 1 28.41 13.89% 0.410 723.14 0.00000
> > 2.6.0-test7 1 28.71 13.92% 0.405 715.02 0.00000
> >
> > With one thread, performance is similar.
> >
> > 2.6.0-test5 8 34.01 17.09% 2.620 834.85 0.00000
> > 2.6.0-test5-mm2 8 32.44 16.44% 2.765 844.99 0.00000
> > 2.6.0-test5-mm4 8 30.75 15.55% 2.967 952.18 0.00000
> > 2.6.0-test6 8 20.67 9.90% 4.477 1569.30 0.00000
> > 2.6.0-test6-mm4 8 21.55 10.31% 4.260 1790.19 0.00000
> > 2.6.0-test7 8 21.46 10.24% 4.312 1913.24 0.00000
> >
> > 2.6.0-test5 16 30.13 15.04% 5.980 1337.69 0.00000
> > 2.6.0-test5-mm2 16 29.71 14.97% 6.024 1115.95 0.00000
> > 2.6.0-test5-mm4 16 27.10 13.65% 6.606 1008.67 0.00000
> > 2.6.0-test6 16 16.44 7.90% 11.201 2820.85 0.00000
> > 2.6.0-test6-mm4 16 16.14 7.78% 11.382 2753.13 0.00000
> > 2.6.0-test7 16 18.34 8.64% 9.877 2742.49 0.00000
> >
> > 2.6.0-test5 32 29.46 14.62% 11.954 1560.02 0.00000
> > 2.6.0-test5-mm2 32 28.01 14.26% 12.808 1627.00 0.00000
> > 2.6.0-test5-mm4 32 28.78 14.63% 12.182 1859.56 0.00000
> > 2.6.0-test6 32 15.60 7.55% 23.415 4877.75 0.00000
> > 2.6.0-test6-mm4 32 14.63 7.20% 24.713 4846.63 0.00000
> > 2.6.0-test7 32 14.79 7.14% 24.234 4855.64 0.00000
> >
> > 2.6.0-test5 64 25.24 13.57% 26.901 1743.05 0.00000
> > 2.6.0-test5-mm2 64 24.01 13.27% 27.922 1849.82 0.00000
> > 2.6.0-test5-mm4 64 23.67 13.15% 28.513 1956.58 0.00000
> > 2.6.0-test6 64 9.45 5.24% 75.491 9578.85 2.75073
> > 2.6.0-test6-mm4 64 8.51 4.82% 84.118 13166.74 3.11647
> > 2.6.0-test7 64 8.58 4.86% 83.430 13622.86 3.05581
> >
> > 2.6.0-test5 128 10.41 9.76% 129.662 1820.78 0.00000
> > 2.6.0-test5-mm2 128 11.00 10.44% 120.118 1742.44 0.00000
> > 2.6.0-test5-mm4 128 10.67 10.15% 122.871 1912.79 0.00000
> > 2.6.0-test6 128 2.97 2.66% 472.251 25694.03 8.39324
> > 2.6.0-test6-mm4 128 2.94 2.66% 473.943 25530.14 8.39028
> > 2.6.0-test7 128 2.94 2.68% 480.175 26306.38 8.39257
> >
> > At 256 threads, throughput is down and max latency skyrockets.
> >
> > 2.6.0-test5 256 9.13 13.30% 305.097 1446.43 0.00000
> > 2.6.0-test5-mm2 256 9.09 13.60% 304.019 1648.80 0.00000
> > 2.6.0-test5-mm4 256 8.63 13.66% 325.793 1542.62 0.00000
> > 2.6.0-test6 256 2.71 2.49% 937.603 529116.60 8.94007
> > 2.6.0-test6-mm4 256 2.62 2.46% 979.221 785760.77 9.12519
> > 2.6.0-test7 256 2.56 2.39% 988.719 582067.86 9.28287
> >
> > Random Reads ext2
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 0.95 0.94% 12.373 114.89 0.00000
> > 2.6.0-test5-mm2 1 0.90 0.88% 13.061 113.19 0.00000
> > 2.6.0-test5-mm4 1 0.53 0.60% 21.903 151.35 0.00000
> > 2.6.0-test6 1 0.73 0.82% 16.053 391.29 0.00000
> > 2.6.0-test6-mm4 1 0.76 0.81% 15.388 126.51 0.00000
> > 2.6.0-test7 1 0.96 0.94% 12.192 101.93 0.00000
> >
> > 2.6.0-test5 8 4.11 4.58% 19.433 137.74 0.00000
> > 2.6.0-test5-mm2 8 3.94 4.03% 20.931 144.16 0.00000
> > 2.6.0-test5-mm4 8 2.81 3.28% 30.086 222.27 0.00000
> > 2.6.0-test6 8 3.25 3.21% 27.598 640.41 0.00000
> > 2.6.0-test6-mm4 8 3.98 3.82% 22.078 324.87 0.00000
> > 2.6.0-test7 8 3.33 3.27% 27.069 417.17 0.00000
> >
> > 2.6.0-test5 16 4.05 3.90% 42.424 236.49 0.00000
> > 2.6.0-test5-mm2 16 4.09 4.31% 41.651 255.17 0.00000
> > 2.6.0-test5-mm4 16 3.05 3.19% 55.847 332.10 0.00000
> > 2.6.0-test6 16 3.43 3.28% 48.836 934.02 0.00000
> > 2.6.0-test6-mm4 16 3.21 2.69% 52.841 936.50 0.00000
> > 2.6.0-test7 16 4.93 4.17% 34.720 630.63 0.00000
> >
> > 2.6.0-test5 32 4.41 3.58% 70.910 516.28 0.00000
> > 2.6.0-test5-mm2 32 4.23 4.17% 78.492 372.63 0.00000
> > 2.6.0-test5-mm4 32 4.19 3.90% 76.001 472.90 0.00000
> > 2.6.0-test6 32 3.74 3.20% 92.407 1578.33 0.00000
> > 2.6.0-test6-mm4 32 2.94 2.97% 116.663 1722.23 0.00000
> > 2.6.0-test7 32 2.62 2.36% 129.093 2248.32 0.00000
> >
> > 2.6.0-test5 64 4.47 4.00% 137.277 1077.53 0.00000
> > 2.6.0-test5-mm2 64 4.79 4.53% 124.035 1144.38 0.00000
> > 2.6.0-test5-mm4 64 4.13 3.81% 143.321 1265.44 0.00000
> > 2.6.0-test6 64 3.89 3.62% 166.555 2679.86 0.00000
> > 2.6.0-test6-mm4 64 3.23 3.15% 200.698 3160.57 0.00000
> > 2.6.0-test7 64 3.14 2.74% 206.849 3384.30 0.00000
> >
> > 2.6.0-test5 128 4.86 4.83% 243.704 2287.09 0.00000
> > 2.6.0-test5-mm2 128 5.08 4.85% 228.102 2391.21 0.00000
> > 2.6.0-test5-mm4 128 4.16 4.65% 270.840 2501.04 0.00000
> > 2.6.0-test6 128 3.19 3.09% 389.583 6092.04 0.12601
> > 2.6.0-test6-mm4 128 2.75 2.78% 439.379 6632.53 0.65524
> > 2.6.0-test7 128 3.16 3.16% 388.762 5810.95 0.07560
> >
> > 2.6.0-test5 256 5.51 7.00% 427.503 2511.79 0.00000
> > 2.6.0-test5-mm2 256 5.23 7.39% 449.007 3169.28 0.00000
> > 2.6.0-test5-mm4 256 4.48 6.76% 526.994 3618.17 0.00000
> > 2.6.0-test6 256 2.01 2.44% 1156.505 14117.02 16.82291
> > 2.6.0-test6-mm4 256 1.99 2.38% 1169.942 13903.51 16.56250
> > 2.6.0-test7 256 2.09 2.41% 1133.624 13195.27 14.97396
> >
> > Sequential and random writes on ext2 are not regressed.
> >
> > ext3 has a regression similar to ext2 for sequential and random reads.
> >
> > Sequential Reads ext3
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 29.03 14.00% 0.401 745.26 0.00000
> > 2.6.0-test5-mm2 1 28.05 13.82% 0.415 763.49 0.00000
> > 2.6.0-test5-mm4 1 29.31 14.10% 0.397 867.89 0.00000
> > 2.6.0-test6 1 28.92 13.83% 0.402 747.59 0.00000
> > 2.6.0-test6-mm4 1 28.01 13.67% 0.415 1162.40 0.00000
> > 2.6.0-test7 1 27.95 13.54% 0.416 860.38 0.00000
> >
> > 2.6.0-test5 8 14.53 8.42% 6.187 1280.16 0.00000
> > 2.6.0-test5-mm2 8 14.25 8.56% 6.428 1438.12 0.00000
> > 2.6.0-test5-mm4 8 14.42 8.60% 6.317 1335.28 0.00000
> > 2.6.0-test6 8 10.37 6.00% 8.791 2065.27 0.00000
> > 2.6.0-test6-mm4 8 9.38 5.81% 9.765 1978.67 0.00000
> > 2.6.0-test7 8 10.25 6.02% 8.931 1977.71 0.00000
> >
> > 2.6.0-test5 16 11.38 7.09% 15.604 1387.68 0.00000
> > 2.6.0-test5-mm2 16 12.84 7.97% 14.206 1246.21 0.00000
> > 2.6.0-test5-mm4 16 12.89 8.00% 14.050 1790.63 0.00000
> > 2.6.0-test6 16 5.13 3.45% 34.925 3982.64 0.00000
> > 2.6.0-test6-mm4 16 4.91 3.42% 37.057 4349.95 0.00000
> > 2.6.0-test7 16 5.27 3.56% 34.774 3852.71 0.00000
> >
> > 2.6.0-test5 32 8.77 6.17% 41.461 1856.08 0.00000
> > 2.6.0-test5-mm2 32 8.79 6.26% 40.822 1667.85 0.00000
> > 2.6.0-test5-mm4 32 8.74 6.41% 41.519 1797.28 0.00000
> > 2.6.0-test6 32 3.32 2.60% 107.252 9267.00 0.00649
> > 2.6.0-test6-mm4 32 3.58 2.78% 100.364 8453.15 0.00496
> > 2.6.0-test7 32 3.39 2.65% 104.283 9029.86 0.00367
> >
> > 2.6.0-test5 64 6.29 5.30% 113.016 1947.14 0.00000
> > 2.6.0-test5-mm2 64 7.58 6.00% 81.717 1722.83 0.00000
> > 2.6.0-test5-mm4 64 7.63 6.07% 79.882 1709.56 0.00000
> > 2.6.0-test6 64 2.63 2.25% 265.280 18736.77 8.76222
> > 2.6.0-test6-mm4 64 2.43 2.17% 277.430 17986.61 8.95066
> > 2.6.0-test7 64 2.40 2.16% 286.947 17890.59 9.59925
> >
> > 2.6.0-test5 128 6.19 6.61% 216.851 1852.30 0.00000
> > 2.6.0-test5-mm2 128 6.73 7.19% 191.761 2049.82 0.00000
> > 2.6.0-test5-mm4 128 6.09 6.55% 209.661 1910.36 0.00000
> > 2.6.0-test6 128 2.06 1.95% 640.506 32660.27 12.41284
> > 2.6.0-test6-mm4 128 2.03 1.98% 645.423 30932.33 12.57601
> > 2.6.0-test7 128 2.05 2.02% 639.791 30885.16 12.72559
> >
> > 2.6.0-test5 256 6.15 9.88% 457.029 1880.47 0.00000
> > 2.6.0-test5-mm2 256 5.51 9.06% 518.943 2002.06 0.00000
> > 2.6.0-test5-mm4 256 5.30 8.81% 535.291 2103.11 0.00000
> > 2.6.0-test6 256 1.85 1.78% 1293.732 230380.55 13.73657
> > 2.6.0-test6-mm4 256 1.91 1.90% 1246.640 243088.10 13.26160
> > 2.6.0-test7 256 1.81 1.78% 1351.071 281394.59 13.80543
> >
> > Random Reads ext3
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 0.90 0.85% 12.963 112.10 0.00000
> > 2.6.0-test5-mm2 1 0.93 0.97% 12.637 125.35 0.00000
> > 2.6.0-test5-mm4 1 0.85 0.95% 13.714 122.66 0.00000
> > 2.6.0-test6 1 0.73 0.71% 16.054 516.87 0.00000
> > 2.6.0-test6-mm4 1 0.91 0.91% 12.907 125.50 0.00000
> > 2.6.0-test7 1 0.94 0.82% 12.443 110.45 0.00000
> >
> > 2.6.0-test5 8 3.78 4.24% 21.477 131.88 0.00000
> > 2.6.0-test5-mm2 8 4.01 4.49% 20.218 124.38 0.00000
> > 2.6.0-test5-mm4 8 4.43 4.63% 18.684 142.49 0.00000
> > 2.6.0-test6 8 3.55 3.11% 23.596 362.51 0.00000
> > 2.6.0-test6-mm4 8 2.52 2.69% 32.927 534.18 0.00000
> > 2.6.0-test7 8 4.21 3.80% 19.881 459.00 0.00000
> >
> > 2.6.0-test5 16 4.10 3.74% 37.249 236.48 0.00000
> > 2.6.0-test5-mm2 16 4.42 4.48% 37.848 208.96 0.00000
> > 2.6.0-test5-mm4 16 4.25 4.51% 36.874 220.75 0.00000
> > 2.6.0-test6 16 3.27 3.13% 49.687 803.39 0.00000
> > 2.6.0-test6-mm4 16 3.37 3.17% 49.112 951.18 0.00000
> > 2.6.0-test7 16 4.49 3.87% 38.468 727.87 0.00000
> >
> > 2.6.0-test5 32 4.28 4.06% 76.377 458.79 0.00000
> > 2.6.0-test5-mm2 32 4.53 4.40% 71.762 466.83 0.00000
> > 2.6.0-test5-mm4 32 4.48 4.23% 70.711 361.46 0.00000
> > 2.6.0-test6 32 3.35 3.06% 98.621 1728.18 0.00000
> > 2.6.0-test6-mm4 32 4.13 3.54% 77.004 1321.89 0.00000
> > 2.6.0-test7 32 3.95 3.30% 79.804 1402.47 0.00000
> >
> > 2.6.0-test5 64 4.89 4.80% 134.964 1243.33 0.00000
> > 2.6.0-test5-mm2 64 4.63 4.17% 108.944 1006.23 0.00000
> > 2.6.0-test5-mm4 64 3.70 3.92% 132.534 1222.82 0.00000
> > 2.6.0-test6 64 3.35 2.98% 189.642 2847.96 0.00000
> > 2.6.0-test6-mm4 64 3.12 3.03% 207.658 3362.27 0.00000
> > 2.6.0-test7 64 3.02 2.89% 212.838 2652.28 0.00000
> >
> > 2.6.0-test5 128 5.26 5.11% 230.823 2088.49 0.00000
> > 2.6.0-test5-mm2 128 4.77 4.65% 252.354 2283.06 0.00000
> > 2.6.0-test5-mm4 128 4.07 4.26% 280.692 2923.29 0.00000
> > 2.6.0-test6 128 2.88 3.07% 431.937 5967.29 0.12601
> > 2.6.0-test6-mm4 128 3.09 3.26% 403.965 5559.31 0.02520
> > 2.6.0-test7 128 3.01 2.95% 407.900 6243.88 0.10081
> >
> > 2.6.0-test5 256 5.18 6.62% 463.528 3270.79 0.00000
> > 2.6.0-test5-mm2 256 5.20 7.12% 452.987 3055.03 0.00000
> > 2.6.0-test5-mm4 256 4.36 5.89% 539.848 3465.46 0.00000
> > 2.6.0-test6 256 2.26 2.92% 1013.076 11495.98 12.47396
> > 2.6.0-test6-mm4 256 2.49 2.98% 943.926 10902.18 10.72917
> > 2.6.0-test7 256 2.43 3.04% 996.420 11016.00 12.08334
> >
> > Sequential Writes ext3
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 56.15 64.55% 0.170 5700.14 0.00010
> > 2.6.0-test5-mm2 1 56.14 67.38% 0.173 4383.06 0.00005
> > 2.6.0-test5-mm4 1 55.64 66.40% 0.173 4323.48 0.00000
> > 2.6.0-test6 1 56.15 65.22% 0.171 4578.10 0.00005
> > 2.6.0-test6-mm4 1 55.81 66.36% 0.172 5623.30 0.00010
> > 2.6.0-test7 1 55.94 65.10% 0.171 4544.88 0.00000
> >
> > 2.6.0-test5 8 2.99 4.94% 23.859 1119399.58 0.08983
> > 2.6.0-test5-mm2 8 3.00 5.08% 24.803 613704.08 0.09155
> > 2.6.0-test5-mm4 8 2.98 5.06% 25.599 515385.61 0.09627
> > 2.6.0-test6 8 4.65 7.53% 15.776 347393.72 0.07853
> > 2.6.0-test6-mm4 8 3.02 5.04% 25.421 399480.58 0.09512
> > 2.6.0-test7 8 3.12 5.12% 23.991 197721.86 0.09255
> >
> > 2.6.0-test5 16 2.58 4.55% 55.284 715963.30 0.14072
> > 2.6.0-test5-mm2 16 2.93 5.09% 50.697 648332.45 0.11883
> > 2.6.0-test5-mm4 16 3.21 5.62% 47.072 544422.32 0.12240
> > 2.6.0-test6 16 3.45 6.12% 39.175 424983.10 0.11596
> > 2.6.0-test6-mm4 16 2.72 4.76% 54.117 429887.30 0.13976
> > 2.6.0-test7 16 2.74 4.73% 52.594 886672.08 0.13404
> >
> > 2.6.0-test5 32 2.28 4.43% 127.746 833586.10 0.20012
> > 2.6.0-test5-mm2 32 2.25 4.58% 124.219 1054925.33 0.20647
> > 2.6.0-test5-mm4 32 2.11 4.33% 139.662 787258.98 0.22011
> > 2.6.0-test6 32 2.80 5.35% 107.223 835055.83 0.18424
> > 2.6.0-test6-mm4 32 2.28 4.26% 117.722 754407.32 0.18582
> > 2.6.0-test7 32 2.21 4.38% 128.475 829790.72 0.21805
> >
> > 2.6.0-test5 64 2.02 4.54% 265.546 1322591.44 0.40007
> > 2.6.0-test5-mm2 64 2.02 4.42% 278.978 1434418.75 0.42186
> > 2.6.0-test5-mm4 64 2.25 4.83% 242.598 1592337.06 0.35062
> > 2.6.0-test6 64 2.44 5.25% 222.436 1233112.15 0.38619
> > 2.6.0-test6-mm4 64 1.98 4.45% 274.091 1442328.98 0.45623
> > 2.6.0-test7 64 2.09 4.37% 249.893 1456365.96 0.30308
> >
> > 2.6.0-test5 128 2.21 4.96% 444.935 2758997.12 0.60168
> > 2.6.0-test5-mm2 128 1.93 4.62% 568.801 3209338.67 0.84395
> > 2.6.0-test5-mm4 128 1.99 4.67% 518.813 2778367.36 0.77595
> > 2.6.0-test6 128 2.00 4.47% 497.636 2933240.88 0.65131
> > 2.6.0-test6-mm4 128 2.03 4.71% 505.696 2839390.84 0.74044
> > 2.6.0-test7 128 2.00 4.69% 491.950 2994053.88 0.75612
> >
> > 2.6.0-test5 256 2.11 7.77% 941.664 4873713.39 1.83378
> > 2.6.0-test5-mm2 256 1.91 6.85% 1093.820 5897453.70 1.80015
> > 2.6.0-test5-mm4 256 1.97 6.74% 1017.367 5695863.45 1.77222
> > 2.6.0-test6 256 1.92 6.71% 1065.957 6104195.68 1.87802
> > 2.6.0-test6-mm4 256 1.93 7.02% 1038.493 5678570.32 1.92114
> > 2.6.0-test7 256 1.85 7.36% 1083.836 5720266.70 1.89524
> >
> > Random Writes ext3
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 3.39 3.44% 0.399 37.37 0.00000
> > 2.6.0-test5-mm2 1 3.22 3.38% 0.516 40.33 0.00000
> > 2.6.0-test5-mm4 1 3.56 3.72% 0.081 10.61 0.00000
> > 2.6.0-test6 1 2.34 2.39% 1.780 24.35 0.00000
> > 2.6.0-test6-mm4 1 3.00 3.19% 0.746 37.62 0.00000
> > 2.6.0-test7 1 3.05 3.06% 0.653 54.78 0.00000
> >
> > 2.6.0-test5 8 3.17 3.59% 13.525 1655.65 0.00000
> > 2.6.0-test5-mm2 8 3.05 3.44% 14.349 2391.37 0.00000
> > 2.6.0-test5-mm4 8 3.08 3.54% 14.093 1401.73 0.00000
> > 2.6.0-test6 8 2.79 3.15% 14.906 3957.84 0.02500
> > 2.6.0-test6-mm4 8 2.88 3.06% 15.029 1283.97 0.00000
> > 2.6.0-test7 8 3.18 3.54% 14.407 1671.73 0.00000
> >
> > 2.6.0-test5 16 3.18 3.49% 26.252 1693.61 0.00000
> > 2.6.0-test5-mm2 16 3.28 3.73% 25.573 3583.93 0.02500
> > 2.6.0-test5-mm4 16 3.21 3.84% 25.857 6012.15 0.02500
> > 2.6.0-test6 16 2.81 2.81% 30.448 2511.41 0.00000
> > 2.6.0-test6-mm4 16 3.17 4.08% 20.815 5062.41 0.02500
> > 2.6.0-test7 16 3.14 3.43% 28.987 2120.35 0.00000
> >
> > 2.6.0-test5 32 3.26 4.05% 46.822 3656.96 0.00000
> > 2.6.0-test5-mm2 32 3.25 4.28% 47.088 2834.74 0.00000
> > 2.6.0-test5-mm4 32 3.29 4.49% 44.765 2409.89 0.00000
> > 2.6.0-test6 32 2.67 2.91% 62.319 11807.07 0.05000
> > 2.6.0-test6-mm4 32 2.98 3.57% 57.025 2250.13 0.00000
> > 2.6.0-test7 32 3.06 3.75% 50.252 9420.48 0.05000
> >
> > 2.6.0-test5 64 3.40 4.96% 46.634 2426.05 0.00000
> > 2.6.0-test5-mm2 64 3.24 4.72% 93.293 7875.05 1.86492
> > 2.6.0-test5-mm4 64 3.24 4.44% 79.998 5292.84 0.63004
> > 2.6.0-test6 64 2.50 3.22% 137.484 6723.01 0.25202
> > 2.6.0-test6-mm4 64 2.83 3.92% 79.495 9204.46 0.68044
> > 2.6.0-test7 64 2.85 4.00% 91.255 6050.08 0.63004
> >
> > 2.6.0-test5 128 3.46 5.89% 101.597 6086.33 0.80645
> > 2.6.0-test5-mm2 128 3.39 5.71% 129.468 6234.78 1.28528
> > 2.6.0-test5-mm4 128 3.27 5.82% 128.846 6601.87 0.65524
> > 2.6.0-test6 128 2.63 4.26% 123.641 6247.05 2.09173
> > 2.6.0-test6-mm4 128 2.69 4.58% 100.787 4669.66 0.00000
> > 2.6.0-test7 128 2.71 4.58% 91.826 4953.82 0.00000
> >
> > 2.6.0-test5 256 3.45 6.88% 210.999 10063.95 6.06770
> > 2.6.0-test5-mm2 256 3.45 6.80% 263.620 8774.35 8.25520
> > 2.6.0-test5-mm4 256 3.33 6.91% 209.373 10921.27 5.10416
> > 2.6.0-test6 256 2.58 5.45% 353.922 7824.45 5.31250
> > 2.6.0-test6-mm4 256 2.99 6.08% 152.068 6246.05 2.36979
> > 2.6.0-test7 256 2.96 6.15% 69.124 3923.92 0.00000
> >
> > Sequential Reads jfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> >
> > 2.6.0-test5 1 28.93 13.16% 0.402 958.45 0.00000
> > 2.6.0-test5-mm2 1 28.84 13.33% 0.403 703.16 0.00000
> > 2.6.0-test5-mm4 1 29.03 13.50% 0.401 871.93 0.00000
> > 2.6.0-test6-mm4 1 28.90 13.54% 0.402 724.31 0.00000
> > 2.6.0-test7 1 29.07 13.44% 0.400 677.09 0.00000
> >
> > 2.6.0-test5 8 50.81 24.01% 1.815 942.91 0.00000
> > 2.6.0-test5-mm2 8 48.22 23.30% 1.918 959.38 0.00000
> > 2.6.0-test5-mm4 8 46.01 21.88% 1.988 840.79 0.00000
> > 2.6.0-test6-mm4 8 20.85 9.38% 4.417 2073.30 0.00000
> > 2.6.0-test7 8 20.88 9.17% 4.402 1941.85 0.00000
> >
> > 2.6.0-test5 16 37.26 17.05% 5.004 855.09 0.00000
> > 2.6.0-test5-mm2 16 40.27 18.83% 4.628 1024.08 0.00000
> > 2.6.0-test5-mm4 16 38.68 18.07% 4.831 820.64 0.00000
> > 2.6.0-test6-mm4 16 11.10 5.05% 16.392 3579.95 0.00000
> > 2.6.0-test7 16 11.98 5.35% 15.258 3381.34 0.00000
> >
> > 2.6.0-test5 32 35.48 16.56% 10.439 1742.70 0.00000
> > 2.6.0-test5-mm2 32 35.46 17.01% 10.292 1979.97 0.00000
> > 2.6.0-test5-mm4 32 33.58 15.89% 11.012 1885.26 0.00000
> > 2.6.0-test6-mm4 32 10.45 4.79% 34.665 6205.82 0.00743
> > 2.6.0-test7 32 10.82 4.85% 33.369 6185.50 0.00668
> >
> > 2.6.0-test5 64 36.58 17.71% 19.855 3297.14 0.00000
> > 2.6.0-test5-mm2 64 34.64 17.14% 21.114 3190.91 0.00000
> > 2.6.0-test5-mm4 64 34.98 17.66% 20.673 3254.24 0.00000
> > 2.6.0-test6-mm4 64 10.64 4.87% 67.501 11935.89 2.07191
> > 2.6.0-test7 64 11.33 5.08% 61.240 11701.91 1.98450
> >
> > 2.6.0-test5 128 11.37 8.64% 109.981 2829.31 0.00000
> > 2.6.0-test5-mm2 128 17.63 11.71% 64.633 2799.09 0.00000
> > 2.6.0-test5-mm4 128 15.92 11.40% 73.376 3032.50 0.00000
> > 2.6.0-test6-mm4 128 2.71 2.52% 520.381 33565.98 8.79664
> > 2.6.0-test7 128 3.09 2.69% 409.208 33147.84 7.51347
> >
> > 2.6.0-test5 256 7.53 11.50% 348.180 2671.65 0.00000
> > 2.6.0-test5-mm2 256 7.77 11.45% 319.292 2285.73 0.00000
> > 2.6.0-test5-mm4 256 7.23 11.94% 364.323 2244.99 0.00000
> > 2.6.0-test6-mm4 256 2.00 1.95% 1023.629 216889.13 12.88433
> > 2.6.0-test7 256 1.85 1.88% 1090.929 455893.85 13.53616
> >
> > Random Reads jfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 0.95 0.82% 12.368 86.79 0.00000
> > 2.6.0-test5-mm2 1 0.97 0.89% 12.013 108.08 0.00000
> > 2.6.0-test5-mm4 1 0.99 0.99% 11.796 110.46 0.00000
> > 2.6.0-test6-mm4 1 0.98 0.91% 11.987 112.94 0.00000
> > 2.6.0-test7 1 0.96 0.83% 12.187 100.56 0.00000
> >
> > 2.6.0-test5 8 4.68 4.23% 18.784 137.13 0.00000
> > 2.6.0-test5-mm2 8 4.06 3.84% 21.942 145.02 0.00000
> > 2.6.0-test5-mm4 8 4.01 3.80% 20.854 150.03 0.00000
> > 2.6.0-test6-mm4 8 0.81 0.61% 112.779 1277.22 0.00000
> > 2.6.0-test7 8 0.82 0.66% 110.357 1250.05 0.00000
> >
> > 2.6.0-test5 16 3.98 3.12% 44.566 188.99 0.00000
> > 2.6.0-test5-mm2 16 4.35 3.42% 40.240 189.97 0.00000
> > 2.6.0-test5-mm4 16 4.28 3.37% 41.846 199.23 0.00000
> > 2.6.0-test6-mm4 16 0.75 0.62% 231.931 2631.93 0.00000
> > 2.6.0-test7 16 0.84 0.65% 212.690 2578.59 0.00000
> >
> > 2.6.0-test5 32 4.09 2.87% 86.440 373.37 0.00000
> > 2.6.0-test5-mm2 32 3.86 3.07% 91.487 443.48 0.00000
> > 2.6.0-test5-mm4 32 3.88 3.07% 90.965 553.03 0.00000
> > 2.6.0-test6-mm4 32 0.76 0.64% 455.886 5230.06 0.00000
> > 2.6.0-test7 32 0.71 0.57% 477.112 5188.75 0.00000
> >
> > 2.6.0-test5 64 4.04 2.88% 169.495 1553.14 0.00000
> > 2.6.0-test5-mm2 64 3.84 3.25% 170.017 1413.19 0.00000
> > 2.6.0-test5-mm4 64 3.99 3.37% 167.645 1628.57 0.00000
> > 2.6.0-test6-mm4 64 0.70 0.61% 950.464 10451.50 26.76412
> > 2.6.0-test7 64 0.71 0.63% 923.148 10352.12 26.38609
> >
> > 2.6.0-test5 128 4.32 3.83% 268.303 2322.59 0.00000
> > 2.6.0-test5-mm2 128 4.18 3.95% 253.072 2451.90 0.00000
> > 2.6.0-test5-mm4 128 4.05 3.95% 255.242 2232.10 0.00000
> > 2.6.0-test6-mm4 128 0.76 0.77% 1626.585 26333.82 35.78629
> > 2.6.0-test7 128 0.78 0.69% 1521.032 24112.21 33.84576
> >
> > 2.6.0-test5 256 4.03 5.43% 585.871 3402.15 0.00000
> > 2.6.0-test5-mm2 256 4.06 5.49% 556.512 3436.30 0.00000
> > 2.6.0-test5-mm4 256 4.11 5.47% 543.649 2806.43 0.00000
> > 2.6.0-test6-mm4 256 0.82 0.87% 2318.070 33523.86 44.84375
> > 2.6.0-test7 256 0.81 0.85% 2315.106 34969.85 44.94791
> >
> > Sequential Writes jfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 56.47 47.37% 0.170 3391.01 0.00000
> > 2.6.0-test5-mm2 1 56.47 49.42% 0.172 2901.46 0.00000
> > 2.6.0-test5-mm4 1 56.40 48.44% 0.181 2578.21 0.00000
> > 2.6.0-test6-mm4 1 56.28 47.99% 0.175 2442.78 0.00000
> > 2.6.0-test7 1 56.39 46.73% 0.171 3405.20 0.00000
> >
> > 2.6.0-test5 8 56.19 65.59% 1.343 8360.10 0.00100
> > 2.6.0-test5-mm2 8 55.94 66.86% 1.336 17088.51 0.00162
> > 2.6.0-test5-mm4 8 56.03 67.52% 1.316 12716.34 0.00105
> > 2.6.0-test6-mm4 8 56.14 65.15% 1.320 13881.60 0.00100
> > 2.6.0-test7 8 56.23 64.40% 1.335 11973.67 0.00109
> >
> > 2.6.0-test5 16 55.15 67.48% 2.651 30109.56 0.03486
> > 2.6.0-test5-mm2 16 55.93 68.92% 2.576 39135.68 0.03081
> > 2.6.0-test5-mm4 16 54.88 69.40% 2.679 36823.40 0.03596
> > 2.6.0-test6-mm4 16 54.85 65.44% 2.641 21368.63 0.04229
> > 2.6.0-test7 16 55.84 68.65% 2.635 63526.06 0.03019
> >
> > 2.6.0-test5 32 54.38 69.36% 5.204 96795.80 0.10614
> > 2.6.0-test5-mm2 32 55.11 70.01% 4.989 55045.13 0.10557
> > 2.6.0-test5-mm4 32 53.92 70.42% 5.099 68154.16 0.10362
> > 2.6.0-test6-mm4 32 54.17 70.60% 5.074 81693.83 0.09809
> > 2.6.0-test7 32 54.75 69.14% 4.998 108095.44 0.09742
> >
> > 2.6.0-test5 64 54.58 75.03% 9.331 208382.29 0.10910
> > 2.6.0-test5-mm2 64 54.44 69.22% 9.195 156878.93 0.11329
> > 2.6.0-test5-mm4 64 53.35 72.64% 9.621 158954.21 0.10639
> > 2.6.0-test6-mm4 64 54.54 70.28% 9.386 148888.99 0.11191
> > 2.6.0-test7 64 54.67 69.68% 8.969 189127.28 0.10605
> >
> > 2.6.0-test5 128 10.39 18.85% 93.764 2075500.73 0.09913
> > 2.6.0-test5-mm2 128 15.87 27.47% 61.244 1306699.59 0.08931
> > 2.6.0-test5-mm4 128 16.06 28.24% 56.744 1437340.48 0.08950
> > 2.6.0-test6-mm4 128 20.89 35.44% 42.971 1018802.82 0.10410
> > 2.6.0-test7 128 13.63 23.07% 69.588 1685679.31 0.08650
> >
> > 2.6.0-test5 256 10.65 20.80% 168.683 1900450.38 0.20290
> > 2.6.0-test5-mm2 256 9.72 18.90% 202.289 2024501.68 0.13385
> > 2.6.0-test5-mm4 256 8.52 17.57% 221.659 1967027.22 0.19260
> > 2.6.0-test6-mm4 256 9.20 17.32% 200.496 1843020.23 0.12737
> > 2.6.0-test7 256 8.85 17.47% 212.298 2080395.84 0.11768
> >
> > Random Writes jfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 3.90 3.48% 0.060 0.56 0.00000
> > 2.6.0-test5-mm2 1 3.89 3.25% 0.061 0.28 0.00000
> > 2.6.0-test5-mm4 1 3.86 3.23% 0.061 0.35 0.00000
> > 2.6.0-test6-mm4 1 3.93 3.40% 0.060 0.37 0.00000
> > 2.6.0-test7 1 3.69 3.03% 0.059 0.88 0.00000
> >
> > 2.6.0-test5 8 3.86 6.54% 0.170 59.53 0.00000
> > 2.6.0-test5-mm2 8 3.68 6.19% 0.170 77.58 0.00000
> > 2.6.0-test5-mm4 8 3.69 6.03% 0.171 69.55 0.00000
> > 2.6.0-test6-mm4 8 3.55 5.93% 0.153 34.88 0.00000
> > 2.6.0-test7 8 3.56 5.99% 0.162 52.74 0.00000
> >
> > 2.6.0-test5 16 3.71 6.47% 0.162 43.67 0.00000
> > 2.6.0-test5-mm2 16 3.82 6.63% 0.166 54.86 0.00000
> > 2.6.0-test5-mm4 16 3.71 6.33% 0.168 43.72 0.00000
> > 2.6.0-test6-mm4 16 3.42 6.01% 0.151 14.21 0.00000
> > 2.6.0-test7 16 3.56 5.94% 0.149 4.41 0.00000
> >
> > 2.6.0-test5 32 3.71 6.48% 0.178 111.32 0.00000
> > 2.6.0-test5-mm2 32 3.50 5.99% 0.148 3.16 0.00000
> > 2.6.0-test5-mm4 32 3.56 6.47% 0.170 55.28 0.00000
> > 2.6.0-test6-mm4 32 3.44 6.09% 0.147 4.23 0.00000
> > 2.6.0-test7 32 3.22 5.41% 0.158 59.51 0.00000
> >
> > 2.6.0-test5 64 3.50 6.28% 0.148 3.96 0.00000
> > 2.6.0-test5-mm2 64 3.34 5.99% 0.160 51.30 0.00000
> > 2.6.0-test5-mm4 64 3.58 6.46% 0.157 7.48 0.00000
> > 2.6.0-test6-mm4 64 3.25 5.67% 0.146 3.61 0.00000
> > 2.6.0-test7 64 3.28 5.74% 0.168 55.67 0.00000
> >
> > 2.6.0-test5 128 3.28 5.72% 20.122 1415.53 0.00000
> > 2.6.0-test5-mm2 128 3.29 5.92% 1.297 1229.30 0.00000
> > 2.6.0-test5-mm4 128 3.42 6.33% 1.499 541.00 0.00000
> > 2.6.0-test6-mm4 128 3.08 5.62% 1.341 1089.07 0.00000
> > 2.6.0-test7 128 3.13 5.51% 0.595 923.55 0.00000
> >
> > 2.6.0-test5 256 2.80 5.18% 89.100 3075.82 0.00000
> > 2.6.0-test5-mm2 256 2.98 5.76% 33.288 2094.84 0.00000
> > 2.6.0-test5-mm4 256 3.19 6.47% 13.738 1648.72 0.00000
> > 2.6.0-test6-mm4 256 2.65 5.49% 18.498 2632.76 0.00000
> > 2.6.0-test7 256 2.74 5.45% 10.049 2382.47 0.00000
> >
> > Sequential Reads reiserfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 28.20 15.37% 0.413 97.43 0.00000
> > 2.6.0-test5-mm2 1 30.21 16.13% 0.385 1159.97 0.00000
> > 2.6.0-test5-mm4 1 28.15 15.29% 0.413 771.52 0.00000
> > 2.6.0-test6 1 28.84 15.63% 0.404 111.79 0.00000
> > 2.6.0-test6-mm4 1 29.99 16.01% 0.388 1065.08 0.00000
> > 2.6.0-test7 1 29.45 15.70% 0.395 113.20 0.00000
> >
> > 2.6.0-test5 8 29.77 17.07% 2.974 854.50 0.00000
> > 2.6.0-test5-mm2 8 31.14 18.03% 2.811 653.61 0.00000
> > 2.6.0-test5-mm4 8 32.23 18.70% 2.727 542.20 0.00000
> > 2.6.0-test6 8 21.41 11.92% 4.275 1487.18 0.00000
> > 2.6.0-test6-mm4 8 21.93 12.30% 4.185 1320.72 0.00000
> > 2.6.0-test7 8 21.40 11.78% 4.263 1261.39 0.00000
> >
> > 2.6.0-test5 16 29.99 17.01% 5.829 901.20 0.00000
> > 2.6.0-test5-mm2 16 29.66 17.05% 5.954 1288.12 0.00000
> > 2.6.0-test5-mm4 16 30.20 17.66% 5.872 1153.05 0.00000
> > 2.6.0-test6 16 19.46 10.99% 9.435 2515.46 0.00000
> > 2.6.0-test6-mm4 16 18.92 10.48% 9.650 2572.79 0.00000
> > 2.6.0-test7 16 18.34 10.11% 10.014 2648.17 0.00000
> >
> > 2.6.0-test5 32 28.86 16.73% 12.003 1408.20 0.00000
> > 2.6.0-test5-mm2 32 28.21 16.52% 12.099 1385.36 0.00000
> > 2.6.0-test5-mm4 32 27.64 16.72% 12.706 1491.53 0.00000
> > 2.6.0-test6 32 17.33 9.87% 20.872 5771.49 0.00005
> > 2.6.0-test6-mm4 32 18.21 10.25% 19.659 4621.02 0.00000
> > 2.6.0-test7 32 17.89 9.88% 20.121 4681.65 0.00000
> >
> > 2.6.0-test5 64 24.08 15.49% 27.936 1610.23 0.00000
> > 2.6.0-test5-mm2 64 23.23 15.01% 27.942 1478.59 0.00000
> > 2.6.0-test5-mm4 64 22.97 15.39% 29.020 1577.31 0.00000
> > 2.6.0-test6 64 7.62 5.28% 95.479 16071.22 3.52058
> > 2.6.0-test6-mm4 64 7.89 5.41% 91.677 12720.86 3.22466
> > 2.6.0-test7 64 7.51 5.16% 96.373 14738.26 3.39680
> >
> > 2.6.0-test5 128 9.77 11.49% 141.347 1501.12 0.00000
> > 2.6.0-test5-mm2 128 10.81 12.13% 121.146 1501.77 0.00000
> > 2.6.0-test5-mm4 128 10.08 12.01% 134.006 1616.24 0.00000
> > 2.6.0-test6 128 2.71 2.80% 534.963 33603.12 10.19945
> > 2.6.0-test6-mm4 128 2.76 2.91% 524.334 34758.05 9.84798
> > 2.6.0-test7 128 2.66 2.72% 543.918 33661.62 10.33516
> >
> > 2.6.0-test5 256 8.88 15.47% 327.283 2378.44 0.00000
> > 2.6.0-test5-mm2 256 8.59 15.37% 339.514 1905.29 0.00000
> > 2.6.0-test5-mm4 256 8.60 16.11% 338.642 1890.15 0.00000
> > 2.6.0-test6 256 2.29 2.42% 1177.401 291951.14 11.55243
> > 2.6.0-test6-mm4 256 2.32 2.50% 1158.512 343146.47 11.20739
> > 2.6.0-test7 256 2.28 2.39% 1172.902 277837.90 11.42778
> >
> > Random Reads reiserfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 0.88 0.95% 13.375 121.48 0.00000
> > 2.6.0-test5-mm2 1 1.01 1.08% 11.543 382.20 0.00000
> > 2.6.0-test5-mm4 1 0.86 1.11% 13.564 123.23 0.00000
> > 2.6.0-test6 1 0.96 1.10% 12.246 127.82 0.00000
> > 2.6.0-test6-mm4 1 1.02 1.16% 11.496 391.26 0.00000
> > 2.6.0-test7 1 0.99 1.04% 11.863 111.96 0.00000
> >
> > 2.6.0-test5 8 4.59 5.28% 17.392 126.60 0.00000
> > 2.6.0-test5-mm2 8 4.88 5.41% 15.978 105.27 0.00000
> > 2.6.0-test5-mm4 8 4.64 5.92% 17.234 119.30 0.00000
> > 2.6.0-test6 8 3.12 3.28% 27.584 373.26 0.00000
> > 2.6.0-test6-mm4 8 3.80 4.12% 23.064 280.06 0.00000
> > 2.6.0-test7 8 3.35 3.69% 24.895 337.23 0.00000
> >
> > 2.6.0-test5 16 4.80 4.33% 34.457 189.80 0.00000
> > 2.6.0-test5-mm2 16 5.27 5.28% 30.882 160.61 0.00000
> > 2.6.0-test5-mm4 16 5.22 5.44% 32.357 494.99 0.00000
> > 2.6.0-test6 16 3.93 4.12% 43.768 728.55 0.00000
> > 2.6.0-test6-mm4 16 3.96 3.79% 41.680 775.69 0.00000
> > 2.6.0-test7 16 3.96 3.85% 42.797 705.06 0.00000
> >
> > 2.6.0-test5 32 5.45 5.07% 59.014 323.30 0.00000
> > 2.6.0-test5-mm2 32 5.65 5.45% 57.673 313.96 0.00000
> > 2.6.0-test5-mm4 32 4.47 4.84% 75.004 455.38 0.00000
> > 2.6.0-test6 32 4.39 4.58% 77.879 1232.67 0.00000
> > 2.6.0-test6-mm4 32 4.15 4.34% 80.355 1229.75 0.00000
> > 2.6.0-test7 32 4.88 4.59% 67.103 1033.43 0.00000
> >
> > 2.6.0-test5 64 5.49 5.42% 108.692 991.66 0.00000
> > 2.6.0-test5-mm2 64 5.10 5.60% 113.940 914.35 0.00000
> > 2.6.0-test5-mm4 64 4.72 5.35% 126.505 1182.30 0.00000
> > 2.6.0-test6 64 3.90 4.03% 165.855 2554.36 0.00000
> > 2.6.0-test6-mm4 64 4.54 4.72% 140.927 2060.27 0.00000
> > 2.6.0-test7 64 4.95 4.73% 129.947 2011.81 0.00000
> >
> > 2.6.0-test5 128 5.48 5.99% 221.089 2074.91 0.00000
> > 2.6.0-test5-mm2 128 5.71 6.49% 208.548 1781.53 0.00000
> > 2.6.0-test5-mm4 128 4.94 5.94% 240.781 2090.92 0.00000
> > 2.6.0-test6 128 3.55 4.55% 367.628 5086.89 0.00000
> > 2.6.0-test6-mm4 128 4.15 4.50% 314.516 3914.52 0.00000
> > 2.6.0-test7 128 3.98 4.56% 328.297 4300.67 0.00000
> >
> > 2.6.0-test5 256 5.39 7.63% 455.466 2448.36 0.00000
> > 2.6.0-test5-mm2 256 5.85 8.92% 415.957 2773.47 0.00000
> > 2.6.0-test5-mm4 256 4.88 8.30% 498.235 2859.83 0.00000
> > 2.6.0-test6 256 2.88 4.24% 861.877 8947.59 7.13542
> > 2.6.0-test6-mm4 256 3.28 4.52% 752.088 7585.14 3.09895
> > 2.6.0-test7 256 3.11 4.19% 794.027 8045.64 3.82812
> >
> > Sequential Writes reiserf
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 55.30 90.83% 0.200 1317.87 0.00000
> > 2.6.0-test5-mm2 1 55.08 92.70% 0.201 1512.52 0.00000
> > 2.6.0-test5-mm4 1 54.62 92.30% 0.202 2146.88 0.00000
> > 2.6.0-test6 1 55.09 91.14% 0.201 1243.21 0.00000
> > 2.6.0-test6-mm4 1 54.46 91.70% 0.202 1725.38 0.00000
> > 2.6.0-test7 1 55.08 91.27% 0.200 2381.55 0.00000
> >
> > 2.6.0-test5 8 27.22 99.99% 2.834 62892.07 0.02932
> > 2.6.0-test5-mm2 8 27.84 100.4% 2.721 20683.58 0.03290
> > 2.6.0-test5-mm4 8 28.18 96.26% 2.667 24253.31 0.03237
> > 2.6.0-test6 8 28.88 102.9% 2.572 28097.64 0.03533
> > 2.6.0-test6-mm4 8 26.57 92.64% 2.855 24274.27 0.05517
> > 2.6.0-test7 8 28.60 104.5% 2.571 19829.03 0.02890
> >
> > 2.6.0-test5 16 30.58 102.6% 4.745 29788.03 0.07806
> > 2.6.0-test5-mm2 16 28.94 101.7% 5.051 37491.58 0.07902
> > 2.6.0-test5-mm4 16 28.45 97.86% 4.971 28952.01 0.08851
> > 2.6.0-test6 16 27.13 98.05% 5.414 58605.69 0.08331
> > 2.6.0-test6-mm4 16 27.04 89.13% 5.347 24565.13 0.08516
> > 2.6.0-test7 16 29.19 98.45% 5.031 29653.75 0.08398
> >
> > 2.6.0-test5 32 27.88 104.6% 10.012 44186.44 0.11630
> > 2.6.0-test5-mm2 32 27.15 104.7% 9.911 47326.17 0.11076
> > 2.6.0-test5-mm4 32 27.57 102.6% 9.820 88173.93 0.10858
> > 2.6.0-test6 32 27.80 100.8% 9.951 72577.05 0.11445
> > 2.6.0-test6-mm4 32 25.80 97.74% 10.563 65400.15 0.11706
> > 2.6.0-test7 32 28.65 105.3% 9.613 44890.58 0.11087
> >
> > 2.6.0-test5 64 27.12 109.5% 19.091 93664.16 0.15073
> > 2.6.0-test5-mm2 64 25.80 117.2% 20.100 73309.25 0.18029
> > 2.6.0-test5-mm4 64 25.64 112.9% 19.408 76139.55 0.16613
> > 2.6.0-test6 64 26.30 112.0% 19.162 93011.66 0.16308
> > 2.6.0-test6-mm4 64 24.15 104.3% 21.613 86562.64 0.18483
> > 2.6.0-test7 64 25.10 107.7% 20.311 80093.32 0.16470
> >
> > 2.6.0-test5 128 23.52 119.4% 42.286 116297.76 0.39005
> > 2.6.0-test5-mm2 128 24.88 127.8% 39.106 139584.40 0.35658
> > 2.6.0-test5-mm4 128 24.28 127.9% 39.111 159921.42 0.37918
> > 2.6.0-test6 128 24.79 122.4% 41.723 127648.00 0.39692
> > 2.6.0-test6-mm4 128 24.48 125.2% 39.716 128936.67 0.38275
> > 2.6.0-test7 128 24.29 122.0% 40.692 152191.29 0.35692
> >
> > 2.6.0-test5 256 19.14 163.4% 112.557 239944.32 1.00599
> > 2.6.0-test5-mm2 256 23.21 153.9% 84.422 200276.08 0.75154
> > 2.6.0-test5-mm4 256 23.45 161.4% 86.423 216520.39 0.71788
> > 2.6.0-test6 256 23.14 168.2% 83.087 247571.24 0.75640
> > 2.6.0-test6-mm4 256 20.87 136.1% 74.033 177291.47 0.66229
> > 2.6.0-test7 256 23.11 142.0% 79.651 233921.91 0.66109
> >
> > Random Writes reiserfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 2.41 2.77% 1.925 282.65 0.00000
> > 2.6.0-test5-mm2 1 3.71 4.05% 0.222 18.72 0.00000
> > 2.6.0-test5-mm4 1 2.29 2.87% 2.191 299.62 0.00000
> > 2.6.0-test6 1 2.89 3.32% 1.040 91.48 0.00000
> > 2.6.0-test6-mm4 1 3.60 3.66% 0.228 24.13 0.00000
> > 2.6.0-test7 1 3.58 3.58% 0.347 30.23 0.00000
> >
> > 2.6.0-test5 8 3.50 4.12% 11.316 1348.95 0.00000
> > 2.6.0-test5-mm2 8 3.74 5.07% 10.135 1133.62 0.00000
> > 2.6.0-test5-mm4 8 3.54 4.73% 10.168 1285.63 0.00000
> > 2.6.0-test6 8 3.26 4.02% 9.287 1182.99 0.00000
> > 2.6.0-test6-mm4 8 3.39 4.53% 5.611 1106.14 0.00000
> > 2.6.0-test7 8 3.29 3.77% 10.518 1191.14 0.00000
> >
> > 2.6.0-test5 16 3.44 4.05% 22.370 4353.76 0.10000
> > 2.6.0-test5-mm2 16 3.54 3.72% 22.545 1478.64 0.00000
> > 2.6.0-test5-mm4 16 3.61 4.71% 20.853 1777.27 0.00000
> > 2.6.0-test6 16 3.60 4.11% 17.724 1421.63 0.00000
> > 2.6.0-test6-mm4 16 3.37 3.72% 23.097 1422.10 0.00000
> > 2.6.0-test7 16 3.46 4.40% 16.822 4822.00 0.02500
> >
> > 2.6.0-test5 32 3.68 4.57% 22.011 1260.71 0.00000
> > 2.6.0-test5-mm2 32 3.60 4.50% 26.561 1836.41 0.00000
> > 2.6.0-test5-mm4 32 3.55 4.32% 33.055 1460.77 0.00000
> > 2.6.0-test6 32 3.41 4.07% 23.498 2013.53 0.00000
> > 2.6.0-test6-mm4 32 3.42 4.36% 13.875 1214.49 0.00000
> > 2.6.0-test7 32 3.42 4.04% 26.600 3437.82 0.02500
> >
> > 2.6.0-test5 64 3.62 5.09% 37.320 8700.40 0.90725
> > 2.6.0-test5-mm2 64 3.81 6.11% 18.225 1814.21 0.00000
> > 2.6.0-test5-mm4 64 3.37 5.09% 43.372 3280.58 0.00000
> > 2.6.0-test6 64 2.96 3.93% 73.537 4484.06 0.00000
> > 2.6.0-test6-mm4 64 3.29 4.98% 22.802 2206.78 0.00000
> > 2.6.0-test7 64 3.42 5.00% 27.782 2453.44 0.00000
> >
> > 2.6.0-test5 128 3.73 5.92% 48.524 7848.01 1.10887
> > 2.6.0-test5-mm2 128 3.39 5.91% 36.168 5651.07 0.25202
> > 2.6.0-test5-mm4 128 3.55 6.74% 70.991 6856.35 1.94052
> > 2.6.0-test6 128 3.16 5.58% 43.277 5286.77 0.05040
> > 2.6.0-test6-mm4 128 3.17 5.25% 43.818 5534.13 0.30242
> > 2.6.0-test7 128 3.26 5.26% 39.051 3037.10 0.00000
> >
> > 2.6.0-test5 256 3.96 8.27% 8.806 963.34 0.00000
> > 2.6.0-test5-mm2 256 3.67 7.17% 20.000 1575.35 0.00000
> > 2.6.0-test5-mm4 256 3.36 6.99% 114.310 4266.84 1.61458
> > 2.6.0-test6 256 3.10 6.69% 132.312 8228.26 3.30729
> > 2.6.0-test6-mm4 256 3.21 6.79% 88.532 8631.04 1.79687
> > 2.6.0-test7 256 3.23 6.61% 48.643 6005.85 0.75521
> >
> > Sequential Reads xfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 30.46 14.80% 0.382 718.97 0.00000
> > 2.6.0-test5-mm2 1 28.43 14.31% 0.409 729.27 0.00000
> > 2.6.0-test5-mm4 1 29.25 14.50% 0.398 751.10 0.00000
> > 2.6.0-test6 1 29.25 14.52% 0.398 735.45 0.00000
> > 2.6.0-test6-mm4 1 28.68 14.60% 0.406 924.52 0.00000
> > 2.6.0-test7 1 29.20 14.46% 0.398 727.21 0.00000
> >
> > 2.6.0-test5 8 55.34 28.06% 1.652 831.66 0.00000
> > 2.6.0-test5-mm2 8 54.19 28.05% 1.707 852.34 0.00000
> > 2.6.0-test5-mm4 8 55.49 28.38% 1.591 1165.42 0.00000
> > 2.6.0-test6 8 25.17 11.78% 3.562 1987.22 0.00000
> > 2.6.0-test6-mm4 8 23.25 11.32% 3.989 1872.56 0.00000
> > 2.6.0-test7 8 25.91 12.19% 3.443 1894.36 0.00000
> >
> > 2.6.0-test5 16 47.04 23.31% 3.903 846.86 0.00000
> > 2.6.0-test5-mm2 16 45.50 23.01% 4.086 1116.23 0.00000
> > 2.6.0-test5-mm4 16 47.00 23.78% 3.827 822.79 0.00000
> > 2.6.0-test6 16 15.07 7.14% 12.090 3088.51 0.00000
> > 2.6.0-test6-mm4 16 12.58 6.23% 14.655 3469.24 0.00000
> > 2.6.0-test7 16 14.79 7.03% 11.808 3178.50 0.00000
> >
> > 2.6.0-test5 32 44.85 22.40% 7.699 1225.63 0.00000
> > 2.6.0-test5-mm2 32 45.89 23.39% 7.624 1173.54 0.00000
> > 2.6.0-test5-mm4 32 44.18 22.68% 7.940 1051.95 0.00000
> > 2.6.0-test6 32 14.80 7.02% 24.021 5758.07 0.00005
> > 2.6.0-test6-mm4 32 14.18 6.98% 25.241 5709.74 0.00000
> > 2.6.0-test7 32 12.46 6.03% 29.086 5836.49 0.00029
> >
> > 2.6.0-test5 64 42.08 21.45% 16.269 1897.20 0.00000
> > 2.6.0-test5-mm2 64 42.36 21.75% 15.948 1925.10 0.00000
> > 2.6.0-test5-mm4 64 43.77 22.73% 15.396 1963.29 0.00000
> > 2.6.0-test6 64 12.50 5.98% 53.864 10887.40 1.74727
> > 2.6.0-test6-mm4 64 12.37 6.19% 55.962 10690.33 1.82176
> > 2.6.0-test7 64 12.25 5.95% 56.211 10899.32 1.78017
> >
> > 2.6.0-test5 128 15.41 12.23% 87.001 2911.16 0.00000
> > 2.6.0-test5-mm2 128 20.45 15.21% 61.549 2516.86 0.00000
> > 2.6.0-test5-mm4 128 18.55 14.38% 68.572 2527.97 0.00000
> > 2.6.0-test6 128 5.20 4.07% 251.782 18918.92 4.34985
> > 2.6.0-test6-mm4 128 4.47 3.56% 290.774 19729.60 5.13649
> > 2.6.0-test7 128 3.50 3.19% 385.879 20057.71 6.42790
> >
> > 2.6.0-test5 256 10.33 13.76% 225.449 6314.51 0.00167
> > 2.6.0-test5-mm2 256 10.58 14.19% 206.025 9145.94 0.00219
> > 2.6.0-test5-mm4 256 11.64 15.40% 175.294 11398.84 0.00677
> > 2.6.0-test6 256 3.84 3.42% 539.606 972621.75 5.75909
> > 2.6.0-test6-mm4 256 3.92 3.35% 527.548 867696.31 5.66021
> > 2.6.0-test7 256 3.99 3.52% 511.822 865743.62 5.56827
> >
> > Random Reads xfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 1.03 0.94% 11.357 107.82 0.00000
> > 2.6.0-test5-mm2 1 1.03 1.02% 11.325 117.87 0.00000
> > 2.6.0-test5-mm4 1 1.04 1.03% 11.261 107.89 0.00000
> > 2.6.0-test6 1 1.03 0.92% 11.386 109.39 0.00000
> > 2.6.0-test6-mm4 1 1.00 1.04% 11.659 119.70 0.00000
> > 2.6.0-test7 1 1.05 0.92% 11.161 96.11 0.00000
> >
> > 2.6.0-test5 8 5.42 5.19% 16.257 124.33 0.00000
> > 2.6.0-test5-mm2 8 5.36 5.79% 16.591 120.16 0.00000
> > 2.6.0-test5-mm4 8 4.67 4.80% 16.448 104.77 0.00000
> > 2.6.0-test6 8 2.08 1.68% 38.061 775.61 0.00000
> > 2.6.0-test6-mm4 8 2.16 1.95% 37.981 686.42 0.00000
> > 2.6.0-test7 8 1.88 1.64% 42.173 683.55 0.00000
> >
> > 2.6.0-test5 16 5.20 4.62% 34.058 146.53 0.00000
> > 2.6.0-test5-mm2 16 5.21 4.88% 34.306 175.88 0.00000
> > 2.6.0-test5-mm4 16 5.18 4.84% 32.751 135.47 0.00000
> > 2.6.0-test6 16 1.86 1.44% 85.898 1244.64 0.00000
> > 2.6.0-test6-mm4 16 1.95 1.64% 80.971 1175.91 0.00000
> > 2.6.0-test7 16 2.35 1.98% 67.271 1001.72 0.00000
> >
> > 2.6.0-test5 32 4.91 3.98% 68.586 265.91 0.00000
> > 2.6.0-test5-mm2 32 5.43 4.74% 63.283 279.95 0.00000
> > 2.6.0-test5-mm4 32 4.69 4.28% 68.452 638.24 0.00000
> > 2.6.0-test6 32 2.92 2.41% 113.627 1626.83 0.00000
> > 2.6.0-test6-mm4 32 3.14 2.92% 105.714 1451.06 0.00000
> > 2.6.0-test7 32 2.43 1.91% 134.936 1864.81 0.00000
> >
> > 2.6.0-test5 64 4.76 4.02% 136.805 879.54 0.00000
> > 2.6.0-test5-mm2 64 4.66 4.68% 136.498 871.90 0.00000
> > 2.6.0-test5-mm4 64 4.81 4.75% 131.041 994.54 0.00000
> > 2.6.0-test6 64 1.45 1.25% 425.661 5262.95 0.02520
> > 2.6.0-test6-mm4 64 1.66 1.40% 382.569 4759.16 0.00000
> > 2.6.0-test7 64 1.41 1.21% 444.027 5820.44 0.30242
> >
> > 2.6.0-test5 128 4.70 4.50% 248.562 1396.77 0.00000
> > 2.6.0-test5-mm2 128 5.29 5.72% 206.313 1395.33 0.00000
> > 2.6.0-test5-mm4 128 5.16 4.95% 209.001 1485.37 0.00000
> > 2.6.0-test6 128 1.65 1.57% 753.959 9498.63 9.02218
> > 2.6.0-test6-mm4 128 1.47 1.43% 833.261 9951.50 11.54233
> > 2.6.0-test7 128 1.23 1.06% 969.453 10793.66 18.04436
> >
> > 2.6.0-test5 256 4.82 6.71% 469.459 2140.97 0.00000
> > 2.6.0-test5-mm2 256 4.60 6.06% 477.682 1934.27 0.00000
> > 2.6.0-test5-mm4 256 5.01 6.60% 402.622 1881.24 0.00000
> > 2.6.0-test6 256 0.90 0.99% 2204.619 32794.62 42.68229
> > 2.6.0-test6-mm4 256 0.88 0.98% 2165.705 33928.64 41.09376
> > 2.6.0-test7 256 0.92 1.05% 2113.981 33637.27 39.11458
> >
> > Sequential Writes xfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 56.44 46.30% 0.182 68379.68 0.00030
> > 2.6.0-test5-mm2 1 56.33 47.39% 0.186 68474.74 0.00030
> > 2.6.0-test5-mm4 1 56.30 46.74% 0.182 68268.99 0.00030
> > 2.6.0-test6 1 56.24 46.08% 0.177 68501.77 0.00030
> > 2.6.0-test6-mm4 1 56.17 47.57% 0.175 68404.17 0.00030
> > 2.6.0-test7 1 56.26 46.03% 0.171 68451.18 0.00030
> >
> > 2.6.0-test5 8 57.54 65.47% 0.991 166706.97 0.00424
> > 2.6.0-test5-mm2 8 57.70 70.44% 1.099 211950.97 0.00420
> > 2.6.0-test5-mm4 8 57.23 66.09% 1.042 205671.61 0.00415
> > 2.6.0-test6 8 57.19 69.31% 1.058 185132.73 0.00410
> > 2.6.0-test6-mm4 8 57.46 66.82% 1.100 178544.18 0.00486
> > 2.6.0-test7 8 57.24 64.93% 1.096 207662.87 0.00481
> >
> > 2.6.0-test5 16 57.26 71.20% 2.069 186551.11 0.01030
> > 2.6.0-test5-mm2 16 57.26 72.02% 2.079 187786.90 0.01059
> > 2.6.0-test5-mm4 16 57.38 68.91% 1.999 193908.75 0.00958
> > 2.6.0-test6 16 57.24 70.72% 2.072 185124.29 0.01092
> > 2.6.0-test6-mm4 16 57.37 68.99% 2.080 208981.25 0.01045
> > 2.6.0-test7 16 57.17 68.79% 2.051 181396.29 0.01135
> >
> > 2.6.0-test5 32 57.29 75.00% 4.076 211670.36 0.01840
> > 2.6.0-test5-mm2 32 57.34 73.21% 3.768 213556.33 0.01560
> > 2.6.0-test5-mm4 32 57.27 75.59% 4.007 199887.88 0.01708
> > 2.6.0-test6 32 57.12 73.89% 4.007 198273.60 0.01750
> > 2.6.0-test6-mm4 32 57.04 73.27% 4.034 181071.17 0.01688
> > 2.6.0-test7 32 57.04 72.85% 4.066 193571.15 0.02007
> >
> > 2.6.0-test5 64 57.13 83.01% 8.186 336901.61 0.03567
> > 2.6.0-test5-mm2 64 56.93 83.63% 8.062 340355.98 0.03066
> > 2.6.0-test5-mm4 64 56.33 82.10% 7.834 325305.64 0.02895
> > 2.6.0-test6 64 56.48 80.10% 7.829 306781.15 0.03157
> > 2.6.0-test6-mm4 64 56.87 81.05% 7.808 257430.27 0.03662
> > 2.6.0-test7 64 56.54 77.01% 7.746 347269.10 0.02880
> >
> > 2.6.0-test5 128 56.58 86.45% 15.451 330514.59 0.06585
> > 2.6.0-test5-mm2 128 56.78 82.31% 14.164 364774.73 0.05298
> > 2.6.0-test5-mm4 128 55.57 82.39% 14.088 350137.61 0.05364
> > 2.6.0-test6 128 56.08 81.22% 13.917 377412.59 0.04669
> > 2.6.0-test6-mm4 128 56.43 79.05% 14.332 357252.17 0.05203
> > 2.6.0-test7 128 56.28 77.14% 14.465 351042.94 0.05546
> >
> > 2.6.0-test5 256 56.13 83.69% 19.836 383235.17 0.07386
> > 2.6.0-test5-mm2 256 56.19 80.90% 16.880 294598.89 0.07434
> > 2.6.0-test5-mm4 256 56.09 87.58% 13.343 308894.90 0.05746
> > 2.6.0-test6 256 55.39 80.57% 18.269 350145.38 0.07200
> > 2.6.0-test6-mm4 256 55.67 81.98% 17.338 337290.04 0.06890
> > 2.6.0-test7 256 55.49 82.29% 15.776 351541.33 0.06071
> >
> > Random Writes xfs
> > Num Avg Maximum Lat%
> > Kernel Thr Rate (CPU%) Latency Latency >2s
> > ------------------------- --- --------------------------------------------
> > 2.6.0-test5 1 3.80 3.61% 0.074 0.46 0.00000
> > 2.6.0-test5-mm2 1 3.86 3.77% 0.077 0.42 0.00000
> > 2.6.0-test5-mm4 1 3.77 3.65% 0.078 1.42 0.00000
> > 2.6.0-test6 1 3.76 3.53% 0.075 1.43 0.00000
> > 2.6.0-test6-mm4 1 3.71 3.64% 0.078 0.68 0.00000
> > 2.6.0-test7 1 3.76 3.53% 0.075 0.87 0.00000
> >
> > 2.6.0-test5 8 3.97 6.78% 0.202 95.59 0.00000
> > 2.6.0-test5-mm2 8 3.91 6.75% 0.181 75.97 0.00000
> > 2.6.0-test5-mm4 8 3.87 6.57% 0.191 81.43 0.00000
> > 2.6.0-test6 8 3.86 6.63% 0.165 25.62 0.00000
> > 2.6.0-test6-mm4 8 3.82 6.33% 0.166 39.50 0.00000
> > 2.6.0-test7 8 3.92 6.44% 0.181 74.95 0.00000
> >
> > 2.6.0-test5 16 3.95 6.81% 0.168 32.20 0.00000
> > 2.6.0-test5-mm2 16 3.94 6.99% 0.175 48.13 0.00000
> > 2.6.0-test5-mm4 16 3.82 6.53% 0.186 87.83 0.00000
> > 2.6.0-test6 16 3.76 6.58% 0.168 25.32 0.00000
> > 2.6.0-test6-mm4 16 3.81 6.62% 0.193 83.97 0.00000
> > 2.6.0-test7 16 3.89 6.57% 0.181 92.60 0.00000
> >
> > 2.6.0-test5 32 3.85 6.64% 0.159 3.38 0.00000
> > 2.6.0-test5-mm2 32 3.80 6.75% 0.175 59.86 0.00000
> > 2.6.0-test5-mm4 32 3.72 6.60% 0.180 55.04 0.00000
> > 2.6.0-test6 32 3.80 6.73% 0.175 56.73 0.00000
> > 2.6.0-test6-mm4 32 3.81 6.61% 0.171 55.65 0.00000
> > 2.6.0-test7 32 3.60 5.97% 0.155 3.09 0.00000
> >
> > 2.6.0-test5 64 3.68 6.56% 0.161 3.90 0.00000
> > 2.6.0-test5-mm2 64 3.60 6.51% 0.163 6.67 0.00000
> > 2.6.0-test5-mm4 64 3.71 6.58% 0.166 5.44 0.00000
> > 2.6.0-test6 64 3.53 6.32% 0.160 4.14 0.00000
> > 2.6.0-test6-mm4 64 3.54 6.21% 0.185 106.22 0.00000
> > 2.6.0-test7 64 3.61 6.15% 0.156 4.34 0.00000
> >
> > 2.6.0-test5 128 3.32 6.23% 0.166 4.42 0.00000
> > 2.6.0-test5-mm2 128 3.64 6.70% 0.163 2.98 0.00000
> > 2.6.0-test5-mm4 128 3.71 6.85% 0.169 5.37 0.00000
> > 2.6.0-test6 128 3.44 6.25% 0.174 63.35 0.00000
> > 2.6.0-test6-mm4 128 3.49 6.36% 0.174 64.77 0.00000
> > 2.6.0-test7 128 3.43 6.15% 0.159 4.05 0.00000
> >
> > 2.6.0-test5 256 3.35 6.56% 0.161 3.65 0.00000
> > 2.6.0-test5-mm2 256 3.29 6.41% 0.164 3.68 0.00000
> > 2.6.0-test5-mm4 256 3.27 6.35% 0.166 4.84 0.00000
> > 2.6.0-test6 256 2.97 5.80% 0.171 47.92 0.00000
> > 2.6.0-test6-mm4 256 3.06 5.93% 0.159 3.02 0.00000
> > 2.6.0-test7 256 3.11 5.88% 0.159 3.91 0.00000
> >
> >
> > dbench did not regress.
> >
> > dbench-1.3 reiserfs 192 processes Average +/- MB/sec
> > 2.6.0-test7 77.3 7.1
> > 2.6.0-test6-mm4 76.6 7.2
> > 2.6.0-test5-mm2 76.1 6.8
> > 2.6.0-test6 73.0 5.1
> > 2.6.0-test5-mm4 72.5 6.6
> > 2.6.0-test5 65.0 1.2
> >
> > dbench-1.3 reiserfs 64 processes Average +/- MB/sec
> > 2.6.0-test5-mm4 111.4 9.1
> > 2.6.0-test6-mm4 109.6 7.5
> > 2.6.0-test6 108.8 4.2
> > 2.6.0-test5-mm2 104.4 8.1
> > 2.6.0-test7 102.3 8.1
> > 2.6.0-test5 100.6 4.7
> >
> > dbench-1.3 xfs 192 processes Average +/- MB/sec
> > 2.6.0-test5 112.5 4.0
> > 2.6.0-test5-mm2 110.8 1.6
> > 2.6.0-test6-mm4 110.7 3.4
> > 2.6.0-test5-mm4 109.7 2.0
> > 2.6.0-test7 109.0 5.1
> > 2.6.0-test6 108.5 6.1
> >
> > dbench-1.3 xfs 64 processes Average +/- MB/sec
> > 2.6.0-test5 143.7 0.9
> > 2.6.0-test7 143.6 2.1
> > 2.6.0-test5-mm2 141.5 1.6
> > 2.6.0-test6 141.4 0.5
> > 2.6.0-test6-mm4 141.2 1.5
> > 2.6.0-test5-mm4 140.0 1.4
> >
> > dbench-1.3 ext2 192 processes Average +/- MB/sec
> > 2.6.0-test6 209.5 11.8
> > 2.6.0-test7 205.2 20.8
> > 2.6.0-test5-mm4 200.8 6.2
> > 2.6.0-test5-mm2 200.0 10.3
> > 2.6.0-test5 198.1 19.5
> > 2.6.0-test6-mm4 193.4 12.0
> >
> > dbench-1.3 ext2 64 processes Average +/- MB/sec
> > 2.6.0-test7 256.8 0.6
> > 2.6.0-test6 256.7 1.6
> > 2.6.0-test5 256.2 1.6
> > 2.6.0-test5-mm4 255.5 1.4
> > 2.6.0-test5-mm2 255.3 0.9
> > 2.6.0-test6-mm4 255.3 1.7
> >
> > dbench-1.3 jfs 192 processes Average +/- MB/sec
> > 2.6.0-test5-mm4 11.3 0.5
> > 2.6.0-test7 11.1 0.2
> > 2.6.0-test6-mm4 11.0 0.3
> >
> > dbench-1.3 jfs 64 processes Average +/- MB/sec
> > 2.6.0-test5-mm2 20.0 2.9
> > 2.6.0-test5 19.5 2.4
> > 2.6.0-test5-mm4 18.6 2.4
> > 2.6.0-test7 18.5 1.1
> > 2.6.0-test6-mm4 18.1 1.6
> >
> > dbench-1.3 ext3 192 processes Average +/- MB/sec
> > 2.6.0-test7 110.8 23.1
> > 2.6.0-test5-mm2 107.5 39.7
> > 2.6.0-test6 103.9 39.0
> > 2.6.0-test6-mm4 99.7 38.7
> > 2.6.0-test5-mm4 82.6 35.5
> > 2.6.0-test5 73.6 15.1
> >
> > dbench-1.3 ext3 64 processes Average +/- MB/sec
> > 2.6.0-test7 177.9 3.4
> > 2.6.0-test6 175.1 3.6
> > 2.6.0-test5 174.5 4.2
> > 2.6.0-test6-mm4 174.2 2.5
> > 2.6.0-test5-mm2 172.8 4.0
> > 2.6.0-test5-mm4 172.6 2.3
> >
> >
> > --
> > Randy Hron
> > http://home.earthlink.net/~rwhron/kernel/bigbox.html
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 1:35 ` Nick Piggin
@ 2003-10-20 4:52 ` Andrew Morton
2003-10-20 8:12 ` Nick Piggin
0 siblings, 1 reply; 25+ messages in thread
From: Andrew Morton @ 2003-10-20 4:52 UTC (permalink / raw)
To: Nick Piggin; +Cc: rwhron, linux-kernel
Nick Piggin <piggin@cyberone.com.au> wrote:
>
>
>
> rwhron@earthlink.net wrote:
>
> >There was about a 50% regression in jobs/minute in AIM7
> >database workload on quad P3 Xeon. The CPU time has not
> >gone up, so the extra run time is coming from something
> >else. (I/O or I/O scheduler?)
> >
> >tiobench sequential reads has a significant regression too.
> >
> >Regression appears unrelated to filesystem type.
> >
> >dbench was not affected.
> >
> >The AIM7 was run on ext2.
> >
>
> Yeah I'd say its all due to the IO scheduler. There is a problem
> I'm thinking about how to fix - its the likely cause of this too.
>
What change do you think it was due to?
It's rather strange that test6 is slow but test6-mm is not: generally the
IO scheduler regressions enter -mm first ;)
Testing versus deadline would be interesting.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 0:37 rwhron
2003-10-20 1:35 ` Nick Piggin
@ 2003-10-20 4:51 ` Dave Olien
2003-10-20 7:55 ` venom
2003-10-24 13:52 ` Magnus Andersson
2 siblings, 1 reply; 25+ messages in thread
From: Dave Olien @ 2003-10-20 4:51 UTC (permalink / raw)
To: rwhron; +Cc: linux-kernel, akpm
Yup, we've seen similar regression on tiobench and reaim workloads.
On Sun, Oct 19, 2003 at 08:37:45PM -0400, rwhron@earthlink.net wrote:
> There was about a 50% regression in jobs/minute in AIM7
> database workload on quad P3 Xeon. The CPU time has not
> gone up, so the extra run time is coming from something
> else. (I/O or I/O scheduler?)
>
> tiobench sequential reads has a significant regression too.
>
> Regression appears unrelated to filesystem type.
>
> dbench was not affected.
>
> The AIM7 was run on ext2.
>
> AIM7 dbase workload
> kernel Tasks Jobs/Min Real CPU
> 2.6.0-test5-mm2 32 609.3 311.9 139.5
> 2.6.0-test5-mm4 32 595.2 319.4 131.7
> 2.6.0-test6-mm4 32 590.8 321.8 130.7
> 2.6.0-test5 32 590.2 322.0 131.9
> 2.6.0-test6 32 320.6 592.9 129.1
> 2.6.0-test7 32 290.3 654.8 120.9
>
> 2.6.0-test5-mm2 64 786.2 483.5 253.2
> 2.6.0-test6-mm4 64 779.4 487.8 253.8
> 2.6.0-test5 64 774.3 490.9 256.6
> 2.6.0-test5-mm4 64 772.0 492.4 253.7
> 2.6.0-test6 64 430.5 883.1 247.6
> 2.6.0-test7 64 402.6 944.2 231.5
>
> 2.6.0-test5-mm2 96 869.2 656.0 382.2
> 2.6.0-test6-mm4 96 867.6 657.2 387.4
> 2.6.0-test5-mm4 96 863.1 660.7 385.2
> 2.6.0-test5 96 859.0 663.9 389.5
> 2.6.0-test6 96 478.5 1191.7 368.1
> 2.6.0-test7 96 453.7 1256.9 346.8
>
> 2.6.0-test5-mm2 128 929.4 818.0 526.1
> 2.6.0-test6-mm4 128 922.6 824.1 523.3
> 2.6.0-test5-mm4 128 916.6 829.5 509.1
> 2.6.0-test5 128 910.3 835.2 528.6
> 2.6.0-test7 128 500.5 1519.1 461.3
> 2.6.0-test6 128 500.2 1520.1 497.4
>
> 2.6.0-test5-mm2 160 962.5 987.4 674.6
> 2.6.0-test5-mm4 160 952.7 997.6 655.5
> 2.6.0-test6-mm4 160 950.0 1000.5 659.5
> 2.6.0-test5 160 741.5 1281.7 644.8
> 2.6.0-test7 160 518.5 1833.1 589.6
> 2.6.0-test6 160 479.6 1981.8 618.9
>
> 2.6.0-test5-mm2 192 993.9 1147.5 810.0
> 2.6.0-test5-mm4 192 982.7 1160.5 788.2
> 2.6.0-test6-mm4 192 978.8 1165.2 808.4
> 2.6.0-test5 192 840.0 1357.8 776.0
> 2.6.0-test7 192 537.9 2120.2 709.3
> 2.6.0-test6 192 494.3 2307.1 748.2
>
> 2.6.0-test5-mm2 224 1011.2 1315.8 943.5
> 2.6.0-test5-mm4 224 1004.7 1324.3 942.1
> 2.6.0-test6-mm4 224 999.5 1331.2 920.0
> 2.6.0-test5 224 986.9 1348.2 911.8
> 2.6.0-test7 224 546.7 2433.8 832.8
> 2.6.0-test6 224 482.9 2755.1 862.1
>
> 2.6.0-test5-mm2 256 1030.9 1475.1 1076.5
> 2.6.0-test5-mm4 256 1027.1 1480.6 1071.4
> 2.6.0-test6-mm4 256 1010.3 1505.2 1074.8
> 2.6.0-test5 256 999.5 1521.3 1029.3
> 2.6.0-test7 256 556.6 2731.9 951.0
> 2.6.0-test6 256 545.1 2789.5 978.4
>
>
> Similarly, AIM7 fileserver benchmark had a drop in throughput.
> CPU time is down, but total runtime is way up.
>
> AIM7 fserver workload
> kernel Tasks Jobs/Min Real CPU
> 2.6.0-test5 4 79.7 304.0 38.3
> 2.6.0-test5-mm2 4 77.4 313.4 38.3
> 2.6.0-test5-mm4 4 76.5 317.1 36.7
> 2.6.0-test6-mm4 4 65.5 369.9 34.5
> 2.6.0-test7 4 45.5 532.5 32.2
> 2.6.0-test6 4 45.0 538.7 31.0
>
> 2.6.0-test5 8 129.8 373.5 66.3
> 2.6.0-test5-mm2 8 103.1 470.4 62.0
> 2.6.0-test5-mm4 8 102.2 474.2 61.8
> 2.6.0-test6 8 82.7 585.9 84.5
> 2.6.0-test6-mm4 8 82.4 588.1 61.1
> 2.6.0-test7 8 61.0 795.0 59.9
>
> 2.6.0-test5 12 164.1 443.0 93.1
> 2.6.0-test5-mm2 12 112.4 646.8 98.4
> 2.6.0-test5-mm4 12 111.2 653.8 96.7
> 2.6.0-test6-mm4 12 95.2 763.6 96.9
> 2.6.0-test6 12 87.9 827.6 118.9
> 2.6.0-test7 12 54.5 1335.0 91.8
>
> 2.6.0-test5 16 173.4 559.0 116.5
> 2.6.0-test5-mm2 16 128.8 752.9 119.9
> 2.6.0-test5-mm4 16 125.4 773.1 124.3
> 2.6.0-test6-mm4 16 104.2 930.3 118.7
> 2.6.0-test6 16 101.4 956.0 107.4
> 2.6.0-test7 16 70.6 1372.6 112.2
>
> 2.6.0-test5 20 180.5 671.5 143.0
> 2.6.0-test5-mm4 20 144.9 836.4 143.0
> 2.6.0-test5-mm2 20 141.7 855.3 144.1
> 2.6.0-test6 20 117.7 1030.0 131.1
> 2.6.0-test6-mm4 20 114.4 1059.5 143.0
> 2.6.0-test7 20 78.7 1539.9 134.0
>
> 2.6.0-test5 24 189.1 769.0 176.0
> 2.6.0-test5-mm2 24 156.2 930.9 180.2
> 2.6.0-test5-mm4 24 154.5 941.6 176.9
> 2.6.0-test6 24 134.5 1081.1 156.5
> 2.6.0-test6-mm4 24 124.5 1167.8 170.2
> 2.6.0-test7 24 76.9 1892.1 169.0
>
> 2.6.0-test5 28 204.6 829.2 197.5
> 2.6.0-test5-mm2 28 158.7 1069.2 209.8
> 2.6.0-test5-mm4 28 156.0 1087.7 211.5
> 2.6.0-test6-mm4 28 134.2 1264.0 196.9
> 2.6.0-test6 28 133.7 1268.7 174.7
> 2.6.0-test7 28 84.1 2016.4 188.5
>
> 2.6.0-test5 32 213.7 907.4 228.9
> 2.6.0-test5-mm2 32 151.9 1276.7 239.7
> 2.6.0-test5-mm4 32 151.5 1279.6 241.4
> 2.6.0-test6 32 147.0 1318.9 197.0
> 2.6.0-test6-mm4 32 138.8 1397.6 225.4
> 2.6.0-test7 32 88.5 2190.3 214.9
>
> Same story for the shared workload. Jobs/minute is down a
> lot since 2.6.0-test5.
>
>
> AIM7 shared workload
> kernel Tasks Jobs/Min Real CPU
> 2.6.0-test5 64 1888.8 197.2 168.3
> 2.6.0-test5-mm2 64 1517.6 245.4 177.0
> 2.6.0-test5-mm4 64 1488.6 250.2 180.5
> 2.6.0-test6-mm4 64 1239.0 300.6 175.0
> 2.6.0-test7 64 1225.1 304.0 166.6
> 2.6.0-test6 64 880.0 423.3 168.1
>
> 2.6.0-test5 128 2094.9 355.6 338.1
> 2.6.0-test5-mm2 128 1837.4 405.4 352.4
> 2.6.0-test5-mm4 128 1826.6 407.8 358.4
> 2.6.0-test6-mm4 128 1561.8 477.0 346.6
> 2.6.0-test7 128 1176.5 633.2 332.6
> 2.6.0-test6 128 1142.1 652.3 332.6
>
> 2.6.0-test5 192 2256.0 495.3 507.8
> 2.6.0-test5-mm2 192 2074.9 538.5 530.9
> 2.6.0-test5-mm4 192 2044.8 546.5 536.9
> 2.6.0-test6-mm4 192 1711.8 652.8 527.3
> 2.6.0-test6 192 1328.3 841.3 505.6
> 2.6.0-test7 192 1317.9 847.9 499.1
>
> 2.6.0-test5 256 2389.0 623.7 683.8
> 2.6.0-test5-mm2 256 2207.2 675.0 720.6
> 2.6.0-test5-mm4 256 2186.2 681.5 729.9
> 2.6.0-test6-mm4 256 1780.7 836.7 719.0
> 2.6.0-test6 256 1479.5 1007.0 672.9
> 2.6.0-test7 256 1282.0 1162.2 672.8
>
> 2.6.0-test5 320 2457.9 757.7 856.6
> 2.6.0-test5-mm4 320 2330.2 799.2 920.0
> 2.6.0-test5-mm2 320 2327.9 800.0 919.4
> 2.6.0-test6-mm4 320 1864.7 998.8 900.9
> 2.6.0-test6 320 1575.9 1181.8 853.5
> 2.6.0-test7 320 1317.9 1413.1 844.4
>
> 2.6.0-test5 384 2480.4 901.0 1036.5
> 2.6.0-test5-mm2 384 2440.2 915.9 1111.6
> 2.6.0-test5-mm4 384 2434.2 918.1 1111.0
> 2.6.0-test6-mm4 384 1883.9 1186.3 1091.7
> 2.6.0-test6 384 1629.8 1371.3 1026.3
> 2.6.0-test7 384 1326.3 1685.0 1018.6
>
> 2.6.0-test5 448 2529.8 1030.7 1209.2
> 2.6.0-test5-mm4 448 2511.6 1038.1 1306.8
> 2.6.0-test5-mm2 448 2510.0 1038.8 1306.1
> 2.6.0-test6-mm4 448 1934.3 1347.9 1283.6
> 2.6.0-test6 448 1612.3 1617.1 1203.3
> 2.6.0-test7 448 1344.7 1939.0 1200.9
>
> 2.6.0-test5-mm2 512 2579.1 1155.4 1504.0
> 2.6.0-test5 512 2566.2 1161.2 1386.2
> 2.6.0-test5-mm4 512 2552.9 1167.2 1509.2
> 2.6.0-test6-mm4 512 1921.5 1550.8 1475.1
> 2.6.0-test6 512 1538.6 1936.7 1376.9
> 2.6.0-test7 512 1370.5 2174.3 1379.2
>
> tiobench read performance in test[67] has a big regression in throughput
> as the number of threads goes up. The max and long latency numbers get
> bad too. It appears to affect ext2, ext3, reiserfs, jfs and xfs.
> Regression is in 2.6.0-test[67] and test6-mm4.
>
> tiobench-0.3.3
> Unit information
> ================
> File size = 8192 megabytes
> Blk Size = 4096 bytes
> Rate = megabytes per second
> CPU% = percentage of CPU used during the test
> Latency = milliseconds
> Lat% = percent of requests that took longer than X seconds
> CPU Eff = Rate divided by CPU% - throughput per cpu load
>
> Sequential Reads ext2
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 28.55 13.89% 0.408 754.59 0.00000
> 2.6.0-test5-mm2 1 28.66 13.98% 0.406 742.76 0.00000
> 2.6.0-test5-mm4 1 28.46 13.90% 0.409 887.02 0.00000
> 2.6.0-test6 1 28.47 13.81% 0.409 1069.74 0.00000
> 2.6.0-test6-mm4 1 28.41 13.89% 0.410 723.14 0.00000
> 2.6.0-test7 1 28.71 13.92% 0.405 715.02 0.00000
>
> With one thread, performance is similar.
>
> 2.6.0-test5 8 34.01 17.09% 2.620 834.85 0.00000
> 2.6.0-test5-mm2 8 32.44 16.44% 2.765 844.99 0.00000
> 2.6.0-test5-mm4 8 30.75 15.55% 2.967 952.18 0.00000
> 2.6.0-test6 8 20.67 9.90% 4.477 1569.30 0.00000
> 2.6.0-test6-mm4 8 21.55 10.31% 4.260 1790.19 0.00000
> 2.6.0-test7 8 21.46 10.24% 4.312 1913.24 0.00000
>
> 2.6.0-test5 16 30.13 15.04% 5.980 1337.69 0.00000
> 2.6.0-test5-mm2 16 29.71 14.97% 6.024 1115.95 0.00000
> 2.6.0-test5-mm4 16 27.10 13.65% 6.606 1008.67 0.00000
> 2.6.0-test6 16 16.44 7.90% 11.201 2820.85 0.00000
> 2.6.0-test6-mm4 16 16.14 7.78% 11.382 2753.13 0.00000
> 2.6.0-test7 16 18.34 8.64% 9.877 2742.49 0.00000
>
> 2.6.0-test5 32 29.46 14.62% 11.954 1560.02 0.00000
> 2.6.0-test5-mm2 32 28.01 14.26% 12.808 1627.00 0.00000
> 2.6.0-test5-mm4 32 28.78 14.63% 12.182 1859.56 0.00000
> 2.6.0-test6 32 15.60 7.55% 23.415 4877.75 0.00000
> 2.6.0-test6-mm4 32 14.63 7.20% 24.713 4846.63 0.00000
> 2.6.0-test7 32 14.79 7.14% 24.234 4855.64 0.00000
>
> 2.6.0-test5 64 25.24 13.57% 26.901 1743.05 0.00000
> 2.6.0-test5-mm2 64 24.01 13.27% 27.922 1849.82 0.00000
> 2.6.0-test5-mm4 64 23.67 13.15% 28.513 1956.58 0.00000
> 2.6.0-test6 64 9.45 5.24% 75.491 9578.85 2.75073
> 2.6.0-test6-mm4 64 8.51 4.82% 84.118 13166.74 3.11647
> 2.6.0-test7 64 8.58 4.86% 83.430 13622.86 3.05581
>
> 2.6.0-test5 128 10.41 9.76% 129.662 1820.78 0.00000
> 2.6.0-test5-mm2 128 11.00 10.44% 120.118 1742.44 0.00000
> 2.6.0-test5-mm4 128 10.67 10.15% 122.871 1912.79 0.00000
> 2.6.0-test6 128 2.97 2.66% 472.251 25694.03 8.39324
> 2.6.0-test6-mm4 128 2.94 2.66% 473.943 25530.14 8.39028
> 2.6.0-test7 128 2.94 2.68% 480.175 26306.38 8.39257
>
> At 256 threads, throughput is down and max latency skyrockets.
>
> 2.6.0-test5 256 9.13 13.30% 305.097 1446.43 0.00000
> 2.6.0-test5-mm2 256 9.09 13.60% 304.019 1648.80 0.00000
> 2.6.0-test5-mm4 256 8.63 13.66% 325.793 1542.62 0.00000
> 2.6.0-test6 256 2.71 2.49% 937.603 529116.60 8.94007
> 2.6.0-test6-mm4 256 2.62 2.46% 979.221 785760.77 9.12519
> 2.6.0-test7 256 2.56 2.39% 988.719 582067.86 9.28287
>
> Random Reads ext2
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 0.95 0.94% 12.373 114.89 0.00000
> 2.6.0-test5-mm2 1 0.90 0.88% 13.061 113.19 0.00000
> 2.6.0-test5-mm4 1 0.53 0.60% 21.903 151.35 0.00000
> 2.6.0-test6 1 0.73 0.82% 16.053 391.29 0.00000
> 2.6.0-test6-mm4 1 0.76 0.81% 15.388 126.51 0.00000
> 2.6.0-test7 1 0.96 0.94% 12.192 101.93 0.00000
>
> 2.6.0-test5 8 4.11 4.58% 19.433 137.74 0.00000
> 2.6.0-test5-mm2 8 3.94 4.03% 20.931 144.16 0.00000
> 2.6.0-test5-mm4 8 2.81 3.28% 30.086 222.27 0.00000
> 2.6.0-test6 8 3.25 3.21% 27.598 640.41 0.00000
> 2.6.0-test6-mm4 8 3.98 3.82% 22.078 324.87 0.00000
> 2.6.0-test7 8 3.33 3.27% 27.069 417.17 0.00000
>
> 2.6.0-test5 16 4.05 3.90% 42.424 236.49 0.00000
> 2.6.0-test5-mm2 16 4.09 4.31% 41.651 255.17 0.00000
> 2.6.0-test5-mm4 16 3.05 3.19% 55.847 332.10 0.00000
> 2.6.0-test6 16 3.43 3.28% 48.836 934.02 0.00000
> 2.6.0-test6-mm4 16 3.21 2.69% 52.841 936.50 0.00000
> 2.6.0-test7 16 4.93 4.17% 34.720 630.63 0.00000
>
> 2.6.0-test5 32 4.41 3.58% 70.910 516.28 0.00000
> 2.6.0-test5-mm2 32 4.23 4.17% 78.492 372.63 0.00000
> 2.6.0-test5-mm4 32 4.19 3.90% 76.001 472.90 0.00000
> 2.6.0-test6 32 3.74 3.20% 92.407 1578.33 0.00000
> 2.6.0-test6-mm4 32 2.94 2.97% 116.663 1722.23 0.00000
> 2.6.0-test7 32 2.62 2.36% 129.093 2248.32 0.00000
>
> 2.6.0-test5 64 4.47 4.00% 137.277 1077.53 0.00000
> 2.6.0-test5-mm2 64 4.79 4.53% 124.035 1144.38 0.00000
> 2.6.0-test5-mm4 64 4.13 3.81% 143.321 1265.44 0.00000
> 2.6.0-test6 64 3.89 3.62% 166.555 2679.86 0.00000
> 2.6.0-test6-mm4 64 3.23 3.15% 200.698 3160.57 0.00000
> 2.6.0-test7 64 3.14 2.74% 206.849 3384.30 0.00000
>
> 2.6.0-test5 128 4.86 4.83% 243.704 2287.09 0.00000
> 2.6.0-test5-mm2 128 5.08 4.85% 228.102 2391.21 0.00000
> 2.6.0-test5-mm4 128 4.16 4.65% 270.840 2501.04 0.00000
> 2.6.0-test6 128 3.19 3.09% 389.583 6092.04 0.12601
> 2.6.0-test6-mm4 128 2.75 2.78% 439.379 6632.53 0.65524
> 2.6.0-test7 128 3.16 3.16% 388.762 5810.95 0.07560
>
> 2.6.0-test5 256 5.51 7.00% 427.503 2511.79 0.00000
> 2.6.0-test5-mm2 256 5.23 7.39% 449.007 3169.28 0.00000
> 2.6.0-test5-mm4 256 4.48 6.76% 526.994 3618.17 0.00000
> 2.6.0-test6 256 2.01 2.44% 1156.505 14117.02 16.82291
> 2.6.0-test6-mm4 256 1.99 2.38% 1169.942 13903.51 16.56250
> 2.6.0-test7 256 2.09 2.41% 1133.624 13195.27 14.97396
>
> Sequential and random writes on ext2 are not regressed.
>
> ext3 has a regression similar to ext2 for sequential and random reads.
>
> Sequential Reads ext3
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 29.03 14.00% 0.401 745.26 0.00000
> 2.6.0-test5-mm2 1 28.05 13.82% 0.415 763.49 0.00000
> 2.6.0-test5-mm4 1 29.31 14.10% 0.397 867.89 0.00000
> 2.6.0-test6 1 28.92 13.83% 0.402 747.59 0.00000
> 2.6.0-test6-mm4 1 28.01 13.67% 0.415 1162.40 0.00000
> 2.6.0-test7 1 27.95 13.54% 0.416 860.38 0.00000
>
> 2.6.0-test5 8 14.53 8.42% 6.187 1280.16 0.00000
> 2.6.0-test5-mm2 8 14.25 8.56% 6.428 1438.12 0.00000
> 2.6.0-test5-mm4 8 14.42 8.60% 6.317 1335.28 0.00000
> 2.6.0-test6 8 10.37 6.00% 8.791 2065.27 0.00000
> 2.6.0-test6-mm4 8 9.38 5.81% 9.765 1978.67 0.00000
> 2.6.0-test7 8 10.25 6.02% 8.931 1977.71 0.00000
>
> 2.6.0-test5 16 11.38 7.09% 15.604 1387.68 0.00000
> 2.6.0-test5-mm2 16 12.84 7.97% 14.206 1246.21 0.00000
> 2.6.0-test5-mm4 16 12.89 8.00% 14.050 1790.63 0.00000
> 2.6.0-test6 16 5.13 3.45% 34.925 3982.64 0.00000
> 2.6.0-test6-mm4 16 4.91 3.42% 37.057 4349.95 0.00000
> 2.6.0-test7 16 5.27 3.56% 34.774 3852.71 0.00000
>
> 2.6.0-test5 32 8.77 6.17% 41.461 1856.08 0.00000
> 2.6.0-test5-mm2 32 8.79 6.26% 40.822 1667.85 0.00000
> 2.6.0-test5-mm4 32 8.74 6.41% 41.519 1797.28 0.00000
> 2.6.0-test6 32 3.32 2.60% 107.252 9267.00 0.00649
> 2.6.0-test6-mm4 32 3.58 2.78% 100.364 8453.15 0.00496
> 2.6.0-test7 32 3.39 2.65% 104.283 9029.86 0.00367
>
> 2.6.0-test5 64 6.29 5.30% 113.016 1947.14 0.00000
> 2.6.0-test5-mm2 64 7.58 6.00% 81.717 1722.83 0.00000
> 2.6.0-test5-mm4 64 7.63 6.07% 79.882 1709.56 0.00000
> 2.6.0-test6 64 2.63 2.25% 265.280 18736.77 8.76222
> 2.6.0-test6-mm4 64 2.43 2.17% 277.430 17986.61 8.95066
> 2.6.0-test7 64 2.40 2.16% 286.947 17890.59 9.59925
>
> 2.6.0-test5 128 6.19 6.61% 216.851 1852.30 0.00000
> 2.6.0-test5-mm2 128 6.73 7.19% 191.761 2049.82 0.00000
> 2.6.0-test5-mm4 128 6.09 6.55% 209.661 1910.36 0.00000
> 2.6.0-test6 128 2.06 1.95% 640.506 32660.27 12.41284
> 2.6.0-test6-mm4 128 2.03 1.98% 645.423 30932.33 12.57601
> 2.6.0-test7 128 2.05 2.02% 639.791 30885.16 12.72559
>
> 2.6.0-test5 256 6.15 9.88% 457.029 1880.47 0.00000
> 2.6.0-test5-mm2 256 5.51 9.06% 518.943 2002.06 0.00000
> 2.6.0-test5-mm4 256 5.30 8.81% 535.291 2103.11 0.00000
> 2.6.0-test6 256 1.85 1.78% 1293.732 230380.55 13.73657
> 2.6.0-test6-mm4 256 1.91 1.90% 1246.640 243088.10 13.26160
> 2.6.0-test7 256 1.81 1.78% 1351.071 281394.59 13.80543
>
> Random Reads ext3
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 0.90 0.85% 12.963 112.10 0.00000
> 2.6.0-test5-mm2 1 0.93 0.97% 12.637 125.35 0.00000
> 2.6.0-test5-mm4 1 0.85 0.95% 13.714 122.66 0.00000
> 2.6.0-test6 1 0.73 0.71% 16.054 516.87 0.00000
> 2.6.0-test6-mm4 1 0.91 0.91% 12.907 125.50 0.00000
> 2.6.0-test7 1 0.94 0.82% 12.443 110.45 0.00000
>
> 2.6.0-test5 8 3.78 4.24% 21.477 131.88 0.00000
> 2.6.0-test5-mm2 8 4.01 4.49% 20.218 124.38 0.00000
> 2.6.0-test5-mm4 8 4.43 4.63% 18.684 142.49 0.00000
> 2.6.0-test6 8 3.55 3.11% 23.596 362.51 0.00000
> 2.6.0-test6-mm4 8 2.52 2.69% 32.927 534.18 0.00000
> 2.6.0-test7 8 4.21 3.80% 19.881 459.00 0.00000
>
> 2.6.0-test5 16 4.10 3.74% 37.249 236.48 0.00000
> 2.6.0-test5-mm2 16 4.42 4.48% 37.848 208.96 0.00000
> 2.6.0-test5-mm4 16 4.25 4.51% 36.874 220.75 0.00000
> 2.6.0-test6 16 3.27 3.13% 49.687 803.39 0.00000
> 2.6.0-test6-mm4 16 3.37 3.17% 49.112 951.18 0.00000
> 2.6.0-test7 16 4.49 3.87% 38.468 727.87 0.00000
>
> 2.6.0-test5 32 4.28 4.06% 76.377 458.79 0.00000
> 2.6.0-test5-mm2 32 4.53 4.40% 71.762 466.83 0.00000
> 2.6.0-test5-mm4 32 4.48 4.23% 70.711 361.46 0.00000
> 2.6.0-test6 32 3.35 3.06% 98.621 1728.18 0.00000
> 2.6.0-test6-mm4 32 4.13 3.54% 77.004 1321.89 0.00000
> 2.6.0-test7 32 3.95 3.30% 79.804 1402.47 0.00000
>
> 2.6.0-test5 64 4.89 4.80% 134.964 1243.33 0.00000
> 2.6.0-test5-mm2 64 4.63 4.17% 108.944 1006.23 0.00000
> 2.6.0-test5-mm4 64 3.70 3.92% 132.534 1222.82 0.00000
> 2.6.0-test6 64 3.35 2.98% 189.642 2847.96 0.00000
> 2.6.0-test6-mm4 64 3.12 3.03% 207.658 3362.27 0.00000
> 2.6.0-test7 64 3.02 2.89% 212.838 2652.28 0.00000
>
> 2.6.0-test5 128 5.26 5.11% 230.823 2088.49 0.00000
> 2.6.0-test5-mm2 128 4.77 4.65% 252.354 2283.06 0.00000
> 2.6.0-test5-mm4 128 4.07 4.26% 280.692 2923.29 0.00000
> 2.6.0-test6 128 2.88 3.07% 431.937 5967.29 0.12601
> 2.6.0-test6-mm4 128 3.09 3.26% 403.965 5559.31 0.02520
> 2.6.0-test7 128 3.01 2.95% 407.900 6243.88 0.10081
>
> 2.6.0-test5 256 5.18 6.62% 463.528 3270.79 0.00000
> 2.6.0-test5-mm2 256 5.20 7.12% 452.987 3055.03 0.00000
> 2.6.0-test5-mm4 256 4.36 5.89% 539.848 3465.46 0.00000
> 2.6.0-test6 256 2.26 2.92% 1013.076 11495.98 12.47396
> 2.6.0-test6-mm4 256 2.49 2.98% 943.926 10902.18 10.72917
> 2.6.0-test7 256 2.43 3.04% 996.420 11016.00 12.08334
>
> Sequential Writes ext3
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 56.15 64.55% 0.170 5700.14 0.00010
> 2.6.0-test5-mm2 1 56.14 67.38% 0.173 4383.06 0.00005
> 2.6.0-test5-mm4 1 55.64 66.40% 0.173 4323.48 0.00000
> 2.6.0-test6 1 56.15 65.22% 0.171 4578.10 0.00005
> 2.6.0-test6-mm4 1 55.81 66.36% 0.172 5623.30 0.00010
> 2.6.0-test7 1 55.94 65.10% 0.171 4544.88 0.00000
>
> 2.6.0-test5 8 2.99 4.94% 23.859 1119399.58 0.08983
> 2.6.0-test5-mm2 8 3.00 5.08% 24.803 613704.08 0.09155
> 2.6.0-test5-mm4 8 2.98 5.06% 25.599 515385.61 0.09627
> 2.6.0-test6 8 4.65 7.53% 15.776 347393.72 0.07853
> 2.6.0-test6-mm4 8 3.02 5.04% 25.421 399480.58 0.09512
> 2.6.0-test7 8 3.12 5.12% 23.991 197721.86 0.09255
>
> 2.6.0-test5 16 2.58 4.55% 55.284 715963.30 0.14072
> 2.6.0-test5-mm2 16 2.93 5.09% 50.697 648332.45 0.11883
> 2.6.0-test5-mm4 16 3.21 5.62% 47.072 544422.32 0.12240
> 2.6.0-test6 16 3.45 6.12% 39.175 424983.10 0.11596
> 2.6.0-test6-mm4 16 2.72 4.76% 54.117 429887.30 0.13976
> 2.6.0-test7 16 2.74 4.73% 52.594 886672.08 0.13404
>
> 2.6.0-test5 32 2.28 4.43% 127.746 833586.10 0.20012
> 2.6.0-test5-mm2 32 2.25 4.58% 124.219 1054925.33 0.20647
> 2.6.0-test5-mm4 32 2.11 4.33% 139.662 787258.98 0.22011
> 2.6.0-test6 32 2.80 5.35% 107.223 835055.83 0.18424
> 2.6.0-test6-mm4 32 2.28 4.26% 117.722 754407.32 0.18582
> 2.6.0-test7 32 2.21 4.38% 128.475 829790.72 0.21805
>
> 2.6.0-test5 64 2.02 4.54% 265.546 1322591.44 0.40007
> 2.6.0-test5-mm2 64 2.02 4.42% 278.978 1434418.75 0.42186
> 2.6.0-test5-mm4 64 2.25 4.83% 242.598 1592337.06 0.35062
> 2.6.0-test6 64 2.44 5.25% 222.436 1233112.15 0.38619
> 2.6.0-test6-mm4 64 1.98 4.45% 274.091 1442328.98 0.45623
> 2.6.0-test7 64 2.09 4.37% 249.893 1456365.96 0.30308
>
> 2.6.0-test5 128 2.21 4.96% 444.935 2758997.12 0.60168
> 2.6.0-test5-mm2 128 1.93 4.62% 568.801 3209338.67 0.84395
> 2.6.0-test5-mm4 128 1.99 4.67% 518.813 2778367.36 0.77595
> 2.6.0-test6 128 2.00 4.47% 497.636 2933240.88 0.65131
> 2.6.0-test6-mm4 128 2.03 4.71% 505.696 2839390.84 0.74044
> 2.6.0-test7 128 2.00 4.69% 491.950 2994053.88 0.75612
>
> 2.6.0-test5 256 2.11 7.77% 941.664 4873713.39 1.83378
> 2.6.0-test5-mm2 256 1.91 6.85% 1093.820 5897453.70 1.80015
> 2.6.0-test5-mm4 256 1.97 6.74% 1017.367 5695863.45 1.77222
> 2.6.0-test6 256 1.92 6.71% 1065.957 6104195.68 1.87802
> 2.6.0-test6-mm4 256 1.93 7.02% 1038.493 5678570.32 1.92114
> 2.6.0-test7 256 1.85 7.36% 1083.836 5720266.70 1.89524
>
> Random Writes ext3
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 3.39 3.44% 0.399 37.37 0.00000
> 2.6.0-test5-mm2 1 3.22 3.38% 0.516 40.33 0.00000
> 2.6.0-test5-mm4 1 3.56 3.72% 0.081 10.61 0.00000
> 2.6.0-test6 1 2.34 2.39% 1.780 24.35 0.00000
> 2.6.0-test6-mm4 1 3.00 3.19% 0.746 37.62 0.00000
> 2.6.0-test7 1 3.05 3.06% 0.653 54.78 0.00000
>
> 2.6.0-test5 8 3.17 3.59% 13.525 1655.65 0.00000
> 2.6.0-test5-mm2 8 3.05 3.44% 14.349 2391.37 0.00000
> 2.6.0-test5-mm4 8 3.08 3.54% 14.093 1401.73 0.00000
> 2.6.0-test6 8 2.79 3.15% 14.906 3957.84 0.02500
> 2.6.0-test6-mm4 8 2.88 3.06% 15.029 1283.97 0.00000
> 2.6.0-test7 8 3.18 3.54% 14.407 1671.73 0.00000
>
> 2.6.0-test5 16 3.18 3.49% 26.252 1693.61 0.00000
> 2.6.0-test5-mm2 16 3.28 3.73% 25.573 3583.93 0.02500
> 2.6.0-test5-mm4 16 3.21 3.84% 25.857 6012.15 0.02500
> 2.6.0-test6 16 2.81 2.81% 30.448 2511.41 0.00000
> 2.6.0-test6-mm4 16 3.17 4.08% 20.815 5062.41 0.02500
> 2.6.0-test7 16 3.14 3.43% 28.987 2120.35 0.00000
>
> 2.6.0-test5 32 3.26 4.05% 46.822 3656.96 0.00000
> 2.6.0-test5-mm2 32 3.25 4.28% 47.088 2834.74 0.00000
> 2.6.0-test5-mm4 32 3.29 4.49% 44.765 2409.89 0.00000
> 2.6.0-test6 32 2.67 2.91% 62.319 11807.07 0.05000
> 2.6.0-test6-mm4 32 2.98 3.57% 57.025 2250.13 0.00000
> 2.6.0-test7 32 3.06 3.75% 50.252 9420.48 0.05000
>
> 2.6.0-test5 64 3.40 4.96% 46.634 2426.05 0.00000
> 2.6.0-test5-mm2 64 3.24 4.72% 93.293 7875.05 1.86492
> 2.6.0-test5-mm4 64 3.24 4.44% 79.998 5292.84 0.63004
> 2.6.0-test6 64 2.50 3.22% 137.484 6723.01 0.25202
> 2.6.0-test6-mm4 64 2.83 3.92% 79.495 9204.46 0.68044
> 2.6.0-test7 64 2.85 4.00% 91.255 6050.08 0.63004
>
> 2.6.0-test5 128 3.46 5.89% 101.597 6086.33 0.80645
> 2.6.0-test5-mm2 128 3.39 5.71% 129.468 6234.78 1.28528
> 2.6.0-test5-mm4 128 3.27 5.82% 128.846 6601.87 0.65524
> 2.6.0-test6 128 2.63 4.26% 123.641 6247.05 2.09173
> 2.6.0-test6-mm4 128 2.69 4.58% 100.787 4669.66 0.00000
> 2.6.0-test7 128 2.71 4.58% 91.826 4953.82 0.00000
>
> 2.6.0-test5 256 3.45 6.88% 210.999 10063.95 6.06770
> 2.6.0-test5-mm2 256 3.45 6.80% 263.620 8774.35 8.25520
> 2.6.0-test5-mm4 256 3.33 6.91% 209.373 10921.27 5.10416
> 2.6.0-test6 256 2.58 5.45% 353.922 7824.45 5.31250
> 2.6.0-test6-mm4 256 2.99 6.08% 152.068 6246.05 2.36979
> 2.6.0-test7 256 2.96 6.15% 69.124 3923.92 0.00000
>
> Sequential Reads jfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
>
> 2.6.0-test5 1 28.93 13.16% 0.402 958.45 0.00000
> 2.6.0-test5-mm2 1 28.84 13.33% 0.403 703.16 0.00000
> 2.6.0-test5-mm4 1 29.03 13.50% 0.401 871.93 0.00000
> 2.6.0-test6-mm4 1 28.90 13.54% 0.402 724.31 0.00000
> 2.6.0-test7 1 29.07 13.44% 0.400 677.09 0.00000
>
> 2.6.0-test5 8 50.81 24.01% 1.815 942.91 0.00000
> 2.6.0-test5-mm2 8 48.22 23.30% 1.918 959.38 0.00000
> 2.6.0-test5-mm4 8 46.01 21.88% 1.988 840.79 0.00000
> 2.6.0-test6-mm4 8 20.85 9.38% 4.417 2073.30 0.00000
> 2.6.0-test7 8 20.88 9.17% 4.402 1941.85 0.00000
>
> 2.6.0-test5 16 37.26 17.05% 5.004 855.09 0.00000
> 2.6.0-test5-mm2 16 40.27 18.83% 4.628 1024.08 0.00000
> 2.6.0-test5-mm4 16 38.68 18.07% 4.831 820.64 0.00000
> 2.6.0-test6-mm4 16 11.10 5.05% 16.392 3579.95 0.00000
> 2.6.0-test7 16 11.98 5.35% 15.258 3381.34 0.00000
>
> 2.6.0-test5 32 35.48 16.56% 10.439 1742.70 0.00000
> 2.6.0-test5-mm2 32 35.46 17.01% 10.292 1979.97 0.00000
> 2.6.0-test5-mm4 32 33.58 15.89% 11.012 1885.26 0.00000
> 2.6.0-test6-mm4 32 10.45 4.79% 34.665 6205.82 0.00743
> 2.6.0-test7 32 10.82 4.85% 33.369 6185.50 0.00668
>
> 2.6.0-test5 64 36.58 17.71% 19.855 3297.14 0.00000
> 2.6.0-test5-mm2 64 34.64 17.14% 21.114 3190.91 0.00000
> 2.6.0-test5-mm4 64 34.98 17.66% 20.673 3254.24 0.00000
> 2.6.0-test6-mm4 64 10.64 4.87% 67.501 11935.89 2.07191
> 2.6.0-test7 64 11.33 5.08% 61.240 11701.91 1.98450
>
> 2.6.0-test5 128 11.37 8.64% 109.981 2829.31 0.00000
> 2.6.0-test5-mm2 128 17.63 11.71% 64.633 2799.09 0.00000
> 2.6.0-test5-mm4 128 15.92 11.40% 73.376 3032.50 0.00000
> 2.6.0-test6-mm4 128 2.71 2.52% 520.381 33565.98 8.79664
> 2.6.0-test7 128 3.09 2.69% 409.208 33147.84 7.51347
>
> 2.6.0-test5 256 7.53 11.50% 348.180 2671.65 0.00000
> 2.6.0-test5-mm2 256 7.77 11.45% 319.292 2285.73 0.00000
> 2.6.0-test5-mm4 256 7.23 11.94% 364.323 2244.99 0.00000
> 2.6.0-test6-mm4 256 2.00 1.95% 1023.629 216889.13 12.88433
> 2.6.0-test7 256 1.85 1.88% 1090.929 455893.85 13.53616
>
> Random Reads jfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 0.95 0.82% 12.368 86.79 0.00000
> 2.6.0-test5-mm2 1 0.97 0.89% 12.013 108.08 0.00000
> 2.6.0-test5-mm4 1 0.99 0.99% 11.796 110.46 0.00000
> 2.6.0-test6-mm4 1 0.98 0.91% 11.987 112.94 0.00000
> 2.6.0-test7 1 0.96 0.83% 12.187 100.56 0.00000
>
> 2.6.0-test5 8 4.68 4.23% 18.784 137.13 0.00000
> 2.6.0-test5-mm2 8 4.06 3.84% 21.942 145.02 0.00000
> 2.6.0-test5-mm4 8 4.01 3.80% 20.854 150.03 0.00000
> 2.6.0-test6-mm4 8 0.81 0.61% 112.779 1277.22 0.00000
> 2.6.0-test7 8 0.82 0.66% 110.357 1250.05 0.00000
>
> 2.6.0-test5 16 3.98 3.12% 44.566 188.99 0.00000
> 2.6.0-test5-mm2 16 4.35 3.42% 40.240 189.97 0.00000
> 2.6.0-test5-mm4 16 4.28 3.37% 41.846 199.23 0.00000
> 2.6.0-test6-mm4 16 0.75 0.62% 231.931 2631.93 0.00000
> 2.6.0-test7 16 0.84 0.65% 212.690 2578.59 0.00000
>
> 2.6.0-test5 32 4.09 2.87% 86.440 373.37 0.00000
> 2.6.0-test5-mm2 32 3.86 3.07% 91.487 443.48 0.00000
> 2.6.0-test5-mm4 32 3.88 3.07% 90.965 553.03 0.00000
> 2.6.0-test6-mm4 32 0.76 0.64% 455.886 5230.06 0.00000
> 2.6.0-test7 32 0.71 0.57% 477.112 5188.75 0.00000
>
> 2.6.0-test5 64 4.04 2.88% 169.495 1553.14 0.00000
> 2.6.0-test5-mm2 64 3.84 3.25% 170.017 1413.19 0.00000
> 2.6.0-test5-mm4 64 3.99 3.37% 167.645 1628.57 0.00000
> 2.6.0-test6-mm4 64 0.70 0.61% 950.464 10451.50 26.76412
> 2.6.0-test7 64 0.71 0.63% 923.148 10352.12 26.38609
>
> 2.6.0-test5 128 4.32 3.83% 268.303 2322.59 0.00000
> 2.6.0-test5-mm2 128 4.18 3.95% 253.072 2451.90 0.00000
> 2.6.0-test5-mm4 128 4.05 3.95% 255.242 2232.10 0.00000
> 2.6.0-test6-mm4 128 0.76 0.77% 1626.585 26333.82 35.78629
> 2.6.0-test7 128 0.78 0.69% 1521.032 24112.21 33.84576
>
> 2.6.0-test5 256 4.03 5.43% 585.871 3402.15 0.00000
> 2.6.0-test5-mm2 256 4.06 5.49% 556.512 3436.30 0.00000
> 2.6.0-test5-mm4 256 4.11 5.47% 543.649 2806.43 0.00000
> 2.6.0-test6-mm4 256 0.82 0.87% 2318.070 33523.86 44.84375
> 2.6.0-test7 256 0.81 0.85% 2315.106 34969.85 44.94791
>
> Sequential Writes jfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 56.47 47.37% 0.170 3391.01 0.00000
> 2.6.0-test5-mm2 1 56.47 49.42% 0.172 2901.46 0.00000
> 2.6.0-test5-mm4 1 56.40 48.44% 0.181 2578.21 0.00000
> 2.6.0-test6-mm4 1 56.28 47.99% 0.175 2442.78 0.00000
> 2.6.0-test7 1 56.39 46.73% 0.171 3405.20 0.00000
>
> 2.6.0-test5 8 56.19 65.59% 1.343 8360.10 0.00100
> 2.6.0-test5-mm2 8 55.94 66.86% 1.336 17088.51 0.00162
> 2.6.0-test5-mm4 8 56.03 67.52% 1.316 12716.34 0.00105
> 2.6.0-test6-mm4 8 56.14 65.15% 1.320 13881.60 0.00100
> 2.6.0-test7 8 56.23 64.40% 1.335 11973.67 0.00109
>
> 2.6.0-test5 16 55.15 67.48% 2.651 30109.56 0.03486
> 2.6.0-test5-mm2 16 55.93 68.92% 2.576 39135.68 0.03081
> 2.6.0-test5-mm4 16 54.88 69.40% 2.679 36823.40 0.03596
> 2.6.0-test6-mm4 16 54.85 65.44% 2.641 21368.63 0.04229
> 2.6.0-test7 16 55.84 68.65% 2.635 63526.06 0.03019
>
> 2.6.0-test5 32 54.38 69.36% 5.204 96795.80 0.10614
> 2.6.0-test5-mm2 32 55.11 70.01% 4.989 55045.13 0.10557
> 2.6.0-test5-mm4 32 53.92 70.42% 5.099 68154.16 0.10362
> 2.6.0-test6-mm4 32 54.17 70.60% 5.074 81693.83 0.09809
> 2.6.0-test7 32 54.75 69.14% 4.998 108095.44 0.09742
>
> 2.6.0-test5 64 54.58 75.03% 9.331 208382.29 0.10910
> 2.6.0-test5-mm2 64 54.44 69.22% 9.195 156878.93 0.11329
> 2.6.0-test5-mm4 64 53.35 72.64% 9.621 158954.21 0.10639
> 2.6.0-test6-mm4 64 54.54 70.28% 9.386 148888.99 0.11191
> 2.6.0-test7 64 54.67 69.68% 8.969 189127.28 0.10605
>
> 2.6.0-test5 128 10.39 18.85% 93.764 2075500.73 0.09913
> 2.6.0-test5-mm2 128 15.87 27.47% 61.244 1306699.59 0.08931
> 2.6.0-test5-mm4 128 16.06 28.24% 56.744 1437340.48 0.08950
> 2.6.0-test6-mm4 128 20.89 35.44% 42.971 1018802.82 0.10410
> 2.6.0-test7 128 13.63 23.07% 69.588 1685679.31 0.08650
>
> 2.6.0-test5 256 10.65 20.80% 168.683 1900450.38 0.20290
> 2.6.0-test5-mm2 256 9.72 18.90% 202.289 2024501.68 0.13385
> 2.6.0-test5-mm4 256 8.52 17.57% 221.659 1967027.22 0.19260
> 2.6.0-test6-mm4 256 9.20 17.32% 200.496 1843020.23 0.12737
> 2.6.0-test7 256 8.85 17.47% 212.298 2080395.84 0.11768
>
> Random Writes jfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 3.90 3.48% 0.060 0.56 0.00000
> 2.6.0-test5-mm2 1 3.89 3.25% 0.061 0.28 0.00000
> 2.6.0-test5-mm4 1 3.86 3.23% 0.061 0.35 0.00000
> 2.6.0-test6-mm4 1 3.93 3.40% 0.060 0.37 0.00000
> 2.6.0-test7 1 3.69 3.03% 0.059 0.88 0.00000
>
> 2.6.0-test5 8 3.86 6.54% 0.170 59.53 0.00000
> 2.6.0-test5-mm2 8 3.68 6.19% 0.170 77.58 0.00000
> 2.6.0-test5-mm4 8 3.69 6.03% 0.171 69.55 0.00000
> 2.6.0-test6-mm4 8 3.55 5.93% 0.153 34.88 0.00000
> 2.6.0-test7 8 3.56 5.99% 0.162 52.74 0.00000
>
> 2.6.0-test5 16 3.71 6.47% 0.162 43.67 0.00000
> 2.6.0-test5-mm2 16 3.82 6.63% 0.166 54.86 0.00000
> 2.6.0-test5-mm4 16 3.71 6.33% 0.168 43.72 0.00000
> 2.6.0-test6-mm4 16 3.42 6.01% 0.151 14.21 0.00000
> 2.6.0-test7 16 3.56 5.94% 0.149 4.41 0.00000
>
> 2.6.0-test5 32 3.71 6.48% 0.178 111.32 0.00000
> 2.6.0-test5-mm2 32 3.50 5.99% 0.148 3.16 0.00000
> 2.6.0-test5-mm4 32 3.56 6.47% 0.170 55.28 0.00000
> 2.6.0-test6-mm4 32 3.44 6.09% 0.147 4.23 0.00000
> 2.6.0-test7 32 3.22 5.41% 0.158 59.51 0.00000
>
> 2.6.0-test5 64 3.50 6.28% 0.148 3.96 0.00000
> 2.6.0-test5-mm2 64 3.34 5.99% 0.160 51.30 0.00000
> 2.6.0-test5-mm4 64 3.58 6.46% 0.157 7.48 0.00000
> 2.6.0-test6-mm4 64 3.25 5.67% 0.146 3.61 0.00000
> 2.6.0-test7 64 3.28 5.74% 0.168 55.67 0.00000
>
> 2.6.0-test5 128 3.28 5.72% 20.122 1415.53 0.00000
> 2.6.0-test5-mm2 128 3.29 5.92% 1.297 1229.30 0.00000
> 2.6.0-test5-mm4 128 3.42 6.33% 1.499 541.00 0.00000
> 2.6.0-test6-mm4 128 3.08 5.62% 1.341 1089.07 0.00000
> 2.6.0-test7 128 3.13 5.51% 0.595 923.55 0.00000
>
> 2.6.0-test5 256 2.80 5.18% 89.100 3075.82 0.00000
> 2.6.0-test5-mm2 256 2.98 5.76% 33.288 2094.84 0.00000
> 2.6.0-test5-mm4 256 3.19 6.47% 13.738 1648.72 0.00000
> 2.6.0-test6-mm4 256 2.65 5.49% 18.498 2632.76 0.00000
> 2.6.0-test7 256 2.74 5.45% 10.049 2382.47 0.00000
>
> Sequential Reads reiserfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 28.20 15.37% 0.413 97.43 0.00000
> 2.6.0-test5-mm2 1 30.21 16.13% 0.385 1159.97 0.00000
> 2.6.0-test5-mm4 1 28.15 15.29% 0.413 771.52 0.00000
> 2.6.0-test6 1 28.84 15.63% 0.404 111.79 0.00000
> 2.6.0-test6-mm4 1 29.99 16.01% 0.388 1065.08 0.00000
> 2.6.0-test7 1 29.45 15.70% 0.395 113.20 0.00000
>
> 2.6.0-test5 8 29.77 17.07% 2.974 854.50 0.00000
> 2.6.0-test5-mm2 8 31.14 18.03% 2.811 653.61 0.00000
> 2.6.0-test5-mm4 8 32.23 18.70% 2.727 542.20 0.00000
> 2.6.0-test6 8 21.41 11.92% 4.275 1487.18 0.00000
> 2.6.0-test6-mm4 8 21.93 12.30% 4.185 1320.72 0.00000
> 2.6.0-test7 8 21.40 11.78% 4.263 1261.39 0.00000
>
> 2.6.0-test5 16 29.99 17.01% 5.829 901.20 0.00000
> 2.6.0-test5-mm2 16 29.66 17.05% 5.954 1288.12 0.00000
> 2.6.0-test5-mm4 16 30.20 17.66% 5.872 1153.05 0.00000
> 2.6.0-test6 16 19.46 10.99% 9.435 2515.46 0.00000
> 2.6.0-test6-mm4 16 18.92 10.48% 9.650 2572.79 0.00000
> 2.6.0-test7 16 18.34 10.11% 10.014 2648.17 0.00000
>
> 2.6.0-test5 32 28.86 16.73% 12.003 1408.20 0.00000
> 2.6.0-test5-mm2 32 28.21 16.52% 12.099 1385.36 0.00000
> 2.6.0-test5-mm4 32 27.64 16.72% 12.706 1491.53 0.00000
> 2.6.0-test6 32 17.33 9.87% 20.872 5771.49 0.00005
> 2.6.0-test6-mm4 32 18.21 10.25% 19.659 4621.02 0.00000
> 2.6.0-test7 32 17.89 9.88% 20.121 4681.65 0.00000
>
> 2.6.0-test5 64 24.08 15.49% 27.936 1610.23 0.00000
> 2.6.0-test5-mm2 64 23.23 15.01% 27.942 1478.59 0.00000
> 2.6.0-test5-mm4 64 22.97 15.39% 29.020 1577.31 0.00000
> 2.6.0-test6 64 7.62 5.28% 95.479 16071.22 3.52058
> 2.6.0-test6-mm4 64 7.89 5.41% 91.677 12720.86 3.22466
> 2.6.0-test7 64 7.51 5.16% 96.373 14738.26 3.39680
>
> 2.6.0-test5 128 9.77 11.49% 141.347 1501.12 0.00000
> 2.6.0-test5-mm2 128 10.81 12.13% 121.146 1501.77 0.00000
> 2.6.0-test5-mm4 128 10.08 12.01% 134.006 1616.24 0.00000
> 2.6.0-test6 128 2.71 2.80% 534.963 33603.12 10.19945
> 2.6.0-test6-mm4 128 2.76 2.91% 524.334 34758.05 9.84798
> 2.6.0-test7 128 2.66 2.72% 543.918 33661.62 10.33516
>
> 2.6.0-test5 256 8.88 15.47% 327.283 2378.44 0.00000
> 2.6.0-test5-mm2 256 8.59 15.37% 339.514 1905.29 0.00000
> 2.6.0-test5-mm4 256 8.60 16.11% 338.642 1890.15 0.00000
> 2.6.0-test6 256 2.29 2.42% 1177.401 291951.14 11.55243
> 2.6.0-test6-mm4 256 2.32 2.50% 1158.512 343146.47 11.20739
> 2.6.0-test7 256 2.28 2.39% 1172.902 277837.90 11.42778
>
> Random Reads reiserfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 0.88 0.95% 13.375 121.48 0.00000
> 2.6.0-test5-mm2 1 1.01 1.08% 11.543 382.20 0.00000
> 2.6.0-test5-mm4 1 0.86 1.11% 13.564 123.23 0.00000
> 2.6.0-test6 1 0.96 1.10% 12.246 127.82 0.00000
> 2.6.0-test6-mm4 1 1.02 1.16% 11.496 391.26 0.00000
> 2.6.0-test7 1 0.99 1.04% 11.863 111.96 0.00000
>
> 2.6.0-test5 8 4.59 5.28% 17.392 126.60 0.00000
> 2.6.0-test5-mm2 8 4.88 5.41% 15.978 105.27 0.00000
> 2.6.0-test5-mm4 8 4.64 5.92% 17.234 119.30 0.00000
> 2.6.0-test6 8 3.12 3.28% 27.584 373.26 0.00000
> 2.6.0-test6-mm4 8 3.80 4.12% 23.064 280.06 0.00000
> 2.6.0-test7 8 3.35 3.69% 24.895 337.23 0.00000
>
> 2.6.0-test5 16 4.80 4.33% 34.457 189.80 0.00000
> 2.6.0-test5-mm2 16 5.27 5.28% 30.882 160.61 0.00000
> 2.6.0-test5-mm4 16 5.22 5.44% 32.357 494.99 0.00000
> 2.6.0-test6 16 3.93 4.12% 43.768 728.55 0.00000
> 2.6.0-test6-mm4 16 3.96 3.79% 41.680 775.69 0.00000
> 2.6.0-test7 16 3.96 3.85% 42.797 705.06 0.00000
>
> 2.6.0-test5 32 5.45 5.07% 59.014 323.30 0.00000
> 2.6.0-test5-mm2 32 5.65 5.45% 57.673 313.96 0.00000
> 2.6.0-test5-mm4 32 4.47 4.84% 75.004 455.38 0.00000
> 2.6.0-test6 32 4.39 4.58% 77.879 1232.67 0.00000
> 2.6.0-test6-mm4 32 4.15 4.34% 80.355 1229.75 0.00000
> 2.6.0-test7 32 4.88 4.59% 67.103 1033.43 0.00000
>
> 2.6.0-test5 64 5.49 5.42% 108.692 991.66 0.00000
> 2.6.0-test5-mm2 64 5.10 5.60% 113.940 914.35 0.00000
> 2.6.0-test5-mm4 64 4.72 5.35% 126.505 1182.30 0.00000
> 2.6.0-test6 64 3.90 4.03% 165.855 2554.36 0.00000
> 2.6.0-test6-mm4 64 4.54 4.72% 140.927 2060.27 0.00000
> 2.6.0-test7 64 4.95 4.73% 129.947 2011.81 0.00000
>
> 2.6.0-test5 128 5.48 5.99% 221.089 2074.91 0.00000
> 2.6.0-test5-mm2 128 5.71 6.49% 208.548 1781.53 0.00000
> 2.6.0-test5-mm4 128 4.94 5.94% 240.781 2090.92 0.00000
> 2.6.0-test6 128 3.55 4.55% 367.628 5086.89 0.00000
> 2.6.0-test6-mm4 128 4.15 4.50% 314.516 3914.52 0.00000
> 2.6.0-test7 128 3.98 4.56% 328.297 4300.67 0.00000
>
> 2.6.0-test5 256 5.39 7.63% 455.466 2448.36 0.00000
> 2.6.0-test5-mm2 256 5.85 8.92% 415.957 2773.47 0.00000
> 2.6.0-test5-mm4 256 4.88 8.30% 498.235 2859.83 0.00000
> 2.6.0-test6 256 2.88 4.24% 861.877 8947.59 7.13542
> 2.6.0-test6-mm4 256 3.28 4.52% 752.088 7585.14 3.09895
> 2.6.0-test7 256 3.11 4.19% 794.027 8045.64 3.82812
>
> Sequential Writes reiserf
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 55.30 90.83% 0.200 1317.87 0.00000
> 2.6.0-test5-mm2 1 55.08 92.70% 0.201 1512.52 0.00000
> 2.6.0-test5-mm4 1 54.62 92.30% 0.202 2146.88 0.00000
> 2.6.0-test6 1 55.09 91.14% 0.201 1243.21 0.00000
> 2.6.0-test6-mm4 1 54.46 91.70% 0.202 1725.38 0.00000
> 2.6.0-test7 1 55.08 91.27% 0.200 2381.55 0.00000
>
> 2.6.0-test5 8 27.22 99.99% 2.834 62892.07 0.02932
> 2.6.0-test5-mm2 8 27.84 100.4% 2.721 20683.58 0.03290
> 2.6.0-test5-mm4 8 28.18 96.26% 2.667 24253.31 0.03237
> 2.6.0-test6 8 28.88 102.9% 2.572 28097.64 0.03533
> 2.6.0-test6-mm4 8 26.57 92.64% 2.855 24274.27 0.05517
> 2.6.0-test7 8 28.60 104.5% 2.571 19829.03 0.02890
>
> 2.6.0-test5 16 30.58 102.6% 4.745 29788.03 0.07806
> 2.6.0-test5-mm2 16 28.94 101.7% 5.051 37491.58 0.07902
> 2.6.0-test5-mm4 16 28.45 97.86% 4.971 28952.01 0.08851
> 2.6.0-test6 16 27.13 98.05% 5.414 58605.69 0.08331
> 2.6.0-test6-mm4 16 27.04 89.13% 5.347 24565.13 0.08516
> 2.6.0-test7 16 29.19 98.45% 5.031 29653.75 0.08398
>
> 2.6.0-test5 32 27.88 104.6% 10.012 44186.44 0.11630
> 2.6.0-test5-mm2 32 27.15 104.7% 9.911 47326.17 0.11076
> 2.6.0-test5-mm4 32 27.57 102.6% 9.820 88173.93 0.10858
> 2.6.0-test6 32 27.80 100.8% 9.951 72577.05 0.11445
> 2.6.0-test6-mm4 32 25.80 97.74% 10.563 65400.15 0.11706
> 2.6.0-test7 32 28.65 105.3% 9.613 44890.58 0.11087
>
> 2.6.0-test5 64 27.12 109.5% 19.091 93664.16 0.15073
> 2.6.0-test5-mm2 64 25.80 117.2% 20.100 73309.25 0.18029
> 2.6.0-test5-mm4 64 25.64 112.9% 19.408 76139.55 0.16613
> 2.6.0-test6 64 26.30 112.0% 19.162 93011.66 0.16308
> 2.6.0-test6-mm4 64 24.15 104.3% 21.613 86562.64 0.18483
> 2.6.0-test7 64 25.10 107.7% 20.311 80093.32 0.16470
>
> 2.6.0-test5 128 23.52 119.4% 42.286 116297.76 0.39005
> 2.6.0-test5-mm2 128 24.88 127.8% 39.106 139584.40 0.35658
> 2.6.0-test5-mm4 128 24.28 127.9% 39.111 159921.42 0.37918
> 2.6.0-test6 128 24.79 122.4% 41.723 127648.00 0.39692
> 2.6.0-test6-mm4 128 24.48 125.2% 39.716 128936.67 0.38275
> 2.6.0-test7 128 24.29 122.0% 40.692 152191.29 0.35692
>
> 2.6.0-test5 256 19.14 163.4% 112.557 239944.32 1.00599
> 2.6.0-test5-mm2 256 23.21 153.9% 84.422 200276.08 0.75154
> 2.6.0-test5-mm4 256 23.45 161.4% 86.423 216520.39 0.71788
> 2.6.0-test6 256 23.14 168.2% 83.087 247571.24 0.75640
> 2.6.0-test6-mm4 256 20.87 136.1% 74.033 177291.47 0.66229
> 2.6.0-test7 256 23.11 142.0% 79.651 233921.91 0.66109
>
> Random Writes reiserfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 2.41 2.77% 1.925 282.65 0.00000
> 2.6.0-test5-mm2 1 3.71 4.05% 0.222 18.72 0.00000
> 2.6.0-test5-mm4 1 2.29 2.87% 2.191 299.62 0.00000
> 2.6.0-test6 1 2.89 3.32% 1.040 91.48 0.00000
> 2.6.0-test6-mm4 1 3.60 3.66% 0.228 24.13 0.00000
> 2.6.0-test7 1 3.58 3.58% 0.347 30.23 0.00000
>
> 2.6.0-test5 8 3.50 4.12% 11.316 1348.95 0.00000
> 2.6.0-test5-mm2 8 3.74 5.07% 10.135 1133.62 0.00000
> 2.6.0-test5-mm4 8 3.54 4.73% 10.168 1285.63 0.00000
> 2.6.0-test6 8 3.26 4.02% 9.287 1182.99 0.00000
> 2.6.0-test6-mm4 8 3.39 4.53% 5.611 1106.14 0.00000
> 2.6.0-test7 8 3.29 3.77% 10.518 1191.14 0.00000
>
> 2.6.0-test5 16 3.44 4.05% 22.370 4353.76 0.10000
> 2.6.0-test5-mm2 16 3.54 3.72% 22.545 1478.64 0.00000
> 2.6.0-test5-mm4 16 3.61 4.71% 20.853 1777.27 0.00000
> 2.6.0-test6 16 3.60 4.11% 17.724 1421.63 0.00000
> 2.6.0-test6-mm4 16 3.37 3.72% 23.097 1422.10 0.00000
> 2.6.0-test7 16 3.46 4.40% 16.822 4822.00 0.02500
>
> 2.6.0-test5 32 3.68 4.57% 22.011 1260.71 0.00000
> 2.6.0-test5-mm2 32 3.60 4.50% 26.561 1836.41 0.00000
> 2.6.0-test5-mm4 32 3.55 4.32% 33.055 1460.77 0.00000
> 2.6.0-test6 32 3.41 4.07% 23.498 2013.53 0.00000
> 2.6.0-test6-mm4 32 3.42 4.36% 13.875 1214.49 0.00000
> 2.6.0-test7 32 3.42 4.04% 26.600 3437.82 0.02500
>
> 2.6.0-test5 64 3.62 5.09% 37.320 8700.40 0.90725
> 2.6.0-test5-mm2 64 3.81 6.11% 18.225 1814.21 0.00000
> 2.6.0-test5-mm4 64 3.37 5.09% 43.372 3280.58 0.00000
> 2.6.0-test6 64 2.96 3.93% 73.537 4484.06 0.00000
> 2.6.0-test6-mm4 64 3.29 4.98% 22.802 2206.78 0.00000
> 2.6.0-test7 64 3.42 5.00% 27.782 2453.44 0.00000
>
> 2.6.0-test5 128 3.73 5.92% 48.524 7848.01 1.10887
> 2.6.0-test5-mm2 128 3.39 5.91% 36.168 5651.07 0.25202
> 2.6.0-test5-mm4 128 3.55 6.74% 70.991 6856.35 1.94052
> 2.6.0-test6 128 3.16 5.58% 43.277 5286.77 0.05040
> 2.6.0-test6-mm4 128 3.17 5.25% 43.818 5534.13 0.30242
> 2.6.0-test7 128 3.26 5.26% 39.051 3037.10 0.00000
>
> 2.6.0-test5 256 3.96 8.27% 8.806 963.34 0.00000
> 2.6.0-test5-mm2 256 3.67 7.17% 20.000 1575.35 0.00000
> 2.6.0-test5-mm4 256 3.36 6.99% 114.310 4266.84 1.61458
> 2.6.0-test6 256 3.10 6.69% 132.312 8228.26 3.30729
> 2.6.0-test6-mm4 256 3.21 6.79% 88.532 8631.04 1.79687
> 2.6.0-test7 256 3.23 6.61% 48.643 6005.85 0.75521
>
> Sequential Reads xfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 30.46 14.80% 0.382 718.97 0.00000
> 2.6.0-test5-mm2 1 28.43 14.31% 0.409 729.27 0.00000
> 2.6.0-test5-mm4 1 29.25 14.50% 0.398 751.10 0.00000
> 2.6.0-test6 1 29.25 14.52% 0.398 735.45 0.00000
> 2.6.0-test6-mm4 1 28.68 14.60% 0.406 924.52 0.00000
> 2.6.0-test7 1 29.20 14.46% 0.398 727.21 0.00000
>
> 2.6.0-test5 8 55.34 28.06% 1.652 831.66 0.00000
> 2.6.0-test5-mm2 8 54.19 28.05% 1.707 852.34 0.00000
> 2.6.0-test5-mm4 8 55.49 28.38% 1.591 1165.42 0.00000
> 2.6.0-test6 8 25.17 11.78% 3.562 1987.22 0.00000
> 2.6.0-test6-mm4 8 23.25 11.32% 3.989 1872.56 0.00000
> 2.6.0-test7 8 25.91 12.19% 3.443 1894.36 0.00000
>
> 2.6.0-test5 16 47.04 23.31% 3.903 846.86 0.00000
> 2.6.0-test5-mm2 16 45.50 23.01% 4.086 1116.23 0.00000
> 2.6.0-test5-mm4 16 47.00 23.78% 3.827 822.79 0.00000
> 2.6.0-test6 16 15.07 7.14% 12.090 3088.51 0.00000
> 2.6.0-test6-mm4 16 12.58 6.23% 14.655 3469.24 0.00000
> 2.6.0-test7 16 14.79 7.03% 11.808 3178.50 0.00000
>
> 2.6.0-test5 32 44.85 22.40% 7.699 1225.63 0.00000
> 2.6.0-test5-mm2 32 45.89 23.39% 7.624 1173.54 0.00000
> 2.6.0-test5-mm4 32 44.18 22.68% 7.940 1051.95 0.00000
> 2.6.0-test6 32 14.80 7.02% 24.021 5758.07 0.00005
> 2.6.0-test6-mm4 32 14.18 6.98% 25.241 5709.74 0.00000
> 2.6.0-test7 32 12.46 6.03% 29.086 5836.49 0.00029
>
> 2.6.0-test5 64 42.08 21.45% 16.269 1897.20 0.00000
> 2.6.0-test5-mm2 64 42.36 21.75% 15.948 1925.10 0.00000
> 2.6.0-test5-mm4 64 43.77 22.73% 15.396 1963.29 0.00000
> 2.6.0-test6 64 12.50 5.98% 53.864 10887.40 1.74727
> 2.6.0-test6-mm4 64 12.37 6.19% 55.962 10690.33 1.82176
> 2.6.0-test7 64 12.25 5.95% 56.211 10899.32 1.78017
>
> 2.6.0-test5 128 15.41 12.23% 87.001 2911.16 0.00000
> 2.6.0-test5-mm2 128 20.45 15.21% 61.549 2516.86 0.00000
> 2.6.0-test5-mm4 128 18.55 14.38% 68.572 2527.97 0.00000
> 2.6.0-test6 128 5.20 4.07% 251.782 18918.92 4.34985
> 2.6.0-test6-mm4 128 4.47 3.56% 290.774 19729.60 5.13649
> 2.6.0-test7 128 3.50 3.19% 385.879 20057.71 6.42790
>
> 2.6.0-test5 256 10.33 13.76% 225.449 6314.51 0.00167
> 2.6.0-test5-mm2 256 10.58 14.19% 206.025 9145.94 0.00219
> 2.6.0-test5-mm4 256 11.64 15.40% 175.294 11398.84 0.00677
> 2.6.0-test6 256 3.84 3.42% 539.606 972621.75 5.75909
> 2.6.0-test6-mm4 256 3.92 3.35% 527.548 867696.31 5.66021
> 2.6.0-test7 256 3.99 3.52% 511.822 865743.62 5.56827
>
> Random Reads xfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 1.03 0.94% 11.357 107.82 0.00000
> 2.6.0-test5-mm2 1 1.03 1.02% 11.325 117.87 0.00000
> 2.6.0-test5-mm4 1 1.04 1.03% 11.261 107.89 0.00000
> 2.6.0-test6 1 1.03 0.92% 11.386 109.39 0.00000
> 2.6.0-test6-mm4 1 1.00 1.04% 11.659 119.70 0.00000
> 2.6.0-test7 1 1.05 0.92% 11.161 96.11 0.00000
>
> 2.6.0-test5 8 5.42 5.19% 16.257 124.33 0.00000
> 2.6.0-test5-mm2 8 5.36 5.79% 16.591 120.16 0.00000
> 2.6.0-test5-mm4 8 4.67 4.80% 16.448 104.77 0.00000
> 2.6.0-test6 8 2.08 1.68% 38.061 775.61 0.00000
> 2.6.0-test6-mm4 8 2.16 1.95% 37.981 686.42 0.00000
> 2.6.0-test7 8 1.88 1.64% 42.173 683.55 0.00000
>
> 2.6.0-test5 16 5.20 4.62% 34.058 146.53 0.00000
> 2.6.0-test5-mm2 16 5.21 4.88% 34.306 175.88 0.00000
> 2.6.0-test5-mm4 16 5.18 4.84% 32.751 135.47 0.00000
> 2.6.0-test6 16 1.86 1.44% 85.898 1244.64 0.00000
> 2.6.0-test6-mm4 16 1.95 1.64% 80.971 1175.91 0.00000
> 2.6.0-test7 16 2.35 1.98% 67.271 1001.72 0.00000
>
> 2.6.0-test5 32 4.91 3.98% 68.586 265.91 0.00000
> 2.6.0-test5-mm2 32 5.43 4.74% 63.283 279.95 0.00000
> 2.6.0-test5-mm4 32 4.69 4.28% 68.452 638.24 0.00000
> 2.6.0-test6 32 2.92 2.41% 113.627 1626.83 0.00000
> 2.6.0-test6-mm4 32 3.14 2.92% 105.714 1451.06 0.00000
> 2.6.0-test7 32 2.43 1.91% 134.936 1864.81 0.00000
>
> 2.6.0-test5 64 4.76 4.02% 136.805 879.54 0.00000
> 2.6.0-test5-mm2 64 4.66 4.68% 136.498 871.90 0.00000
> 2.6.0-test5-mm4 64 4.81 4.75% 131.041 994.54 0.00000
> 2.6.0-test6 64 1.45 1.25% 425.661 5262.95 0.02520
> 2.6.0-test6-mm4 64 1.66 1.40% 382.569 4759.16 0.00000
> 2.6.0-test7 64 1.41 1.21% 444.027 5820.44 0.30242
>
> 2.6.0-test5 128 4.70 4.50% 248.562 1396.77 0.00000
> 2.6.0-test5-mm2 128 5.29 5.72% 206.313 1395.33 0.00000
> 2.6.0-test5-mm4 128 5.16 4.95% 209.001 1485.37 0.00000
> 2.6.0-test6 128 1.65 1.57% 753.959 9498.63 9.02218
> 2.6.0-test6-mm4 128 1.47 1.43% 833.261 9951.50 11.54233
> 2.6.0-test7 128 1.23 1.06% 969.453 10793.66 18.04436
>
> 2.6.0-test5 256 4.82 6.71% 469.459 2140.97 0.00000
> 2.6.0-test5-mm2 256 4.60 6.06% 477.682 1934.27 0.00000
> 2.6.0-test5-mm4 256 5.01 6.60% 402.622 1881.24 0.00000
> 2.6.0-test6 256 0.90 0.99% 2204.619 32794.62 42.68229
> 2.6.0-test6-mm4 256 0.88 0.98% 2165.705 33928.64 41.09376
> 2.6.0-test7 256 0.92 1.05% 2113.981 33637.27 39.11458
>
> Sequential Writes xfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 56.44 46.30% 0.182 68379.68 0.00030
> 2.6.0-test5-mm2 1 56.33 47.39% 0.186 68474.74 0.00030
> 2.6.0-test5-mm4 1 56.30 46.74% 0.182 68268.99 0.00030
> 2.6.0-test6 1 56.24 46.08% 0.177 68501.77 0.00030
> 2.6.0-test6-mm4 1 56.17 47.57% 0.175 68404.17 0.00030
> 2.6.0-test7 1 56.26 46.03% 0.171 68451.18 0.00030
>
> 2.6.0-test5 8 57.54 65.47% 0.991 166706.97 0.00424
> 2.6.0-test5-mm2 8 57.70 70.44% 1.099 211950.97 0.00420
> 2.6.0-test5-mm4 8 57.23 66.09% 1.042 205671.61 0.00415
> 2.6.0-test6 8 57.19 69.31% 1.058 185132.73 0.00410
> 2.6.0-test6-mm4 8 57.46 66.82% 1.100 178544.18 0.00486
> 2.6.0-test7 8 57.24 64.93% 1.096 207662.87 0.00481
>
> 2.6.0-test5 16 57.26 71.20% 2.069 186551.11 0.01030
> 2.6.0-test5-mm2 16 57.26 72.02% 2.079 187786.90 0.01059
> 2.6.0-test5-mm4 16 57.38 68.91% 1.999 193908.75 0.00958
> 2.6.0-test6 16 57.24 70.72% 2.072 185124.29 0.01092
> 2.6.0-test6-mm4 16 57.37 68.99% 2.080 208981.25 0.01045
> 2.6.0-test7 16 57.17 68.79% 2.051 181396.29 0.01135
>
> 2.6.0-test5 32 57.29 75.00% 4.076 211670.36 0.01840
> 2.6.0-test5-mm2 32 57.34 73.21% 3.768 213556.33 0.01560
> 2.6.0-test5-mm4 32 57.27 75.59% 4.007 199887.88 0.01708
> 2.6.0-test6 32 57.12 73.89% 4.007 198273.60 0.01750
> 2.6.0-test6-mm4 32 57.04 73.27% 4.034 181071.17 0.01688
> 2.6.0-test7 32 57.04 72.85% 4.066 193571.15 0.02007
>
> 2.6.0-test5 64 57.13 83.01% 8.186 336901.61 0.03567
> 2.6.0-test5-mm2 64 56.93 83.63% 8.062 340355.98 0.03066
> 2.6.0-test5-mm4 64 56.33 82.10% 7.834 325305.64 0.02895
> 2.6.0-test6 64 56.48 80.10% 7.829 306781.15 0.03157
> 2.6.0-test6-mm4 64 56.87 81.05% 7.808 257430.27 0.03662
> 2.6.0-test7 64 56.54 77.01% 7.746 347269.10 0.02880
>
> 2.6.0-test5 128 56.58 86.45% 15.451 330514.59 0.06585
> 2.6.0-test5-mm2 128 56.78 82.31% 14.164 364774.73 0.05298
> 2.6.0-test5-mm4 128 55.57 82.39% 14.088 350137.61 0.05364
> 2.6.0-test6 128 56.08 81.22% 13.917 377412.59 0.04669
> 2.6.0-test6-mm4 128 56.43 79.05% 14.332 357252.17 0.05203
> 2.6.0-test7 128 56.28 77.14% 14.465 351042.94 0.05546
>
> 2.6.0-test5 256 56.13 83.69% 19.836 383235.17 0.07386
> 2.6.0-test5-mm2 256 56.19 80.90% 16.880 294598.89 0.07434
> 2.6.0-test5-mm4 256 56.09 87.58% 13.343 308894.90 0.05746
> 2.6.0-test6 256 55.39 80.57% 18.269 350145.38 0.07200
> 2.6.0-test6-mm4 256 55.67 81.98% 17.338 337290.04 0.06890
> 2.6.0-test7 256 55.49 82.29% 15.776 351541.33 0.06071
>
> Random Writes xfs
> Num Avg Maximum Lat%
> Kernel Thr Rate (CPU%) Latency Latency >2s
> ------------------------- --- --------------------------------------------
> 2.6.0-test5 1 3.80 3.61% 0.074 0.46 0.00000
> 2.6.0-test5-mm2 1 3.86 3.77% 0.077 0.42 0.00000
> 2.6.0-test5-mm4 1 3.77 3.65% 0.078 1.42 0.00000
> 2.6.0-test6 1 3.76 3.53% 0.075 1.43 0.00000
> 2.6.0-test6-mm4 1 3.71 3.64% 0.078 0.68 0.00000
> 2.6.0-test7 1 3.76 3.53% 0.075 0.87 0.00000
>
> 2.6.0-test5 8 3.97 6.78% 0.202 95.59 0.00000
> 2.6.0-test5-mm2 8 3.91 6.75% 0.181 75.97 0.00000
> 2.6.0-test5-mm4 8 3.87 6.57% 0.191 81.43 0.00000
> 2.6.0-test6 8 3.86 6.63% 0.165 25.62 0.00000
> 2.6.0-test6-mm4 8 3.82 6.33% 0.166 39.50 0.00000
> 2.6.0-test7 8 3.92 6.44% 0.181 74.95 0.00000
>
> 2.6.0-test5 16 3.95 6.81% 0.168 32.20 0.00000
> 2.6.0-test5-mm2 16 3.94 6.99% 0.175 48.13 0.00000
> 2.6.0-test5-mm4 16 3.82 6.53% 0.186 87.83 0.00000
> 2.6.0-test6 16 3.76 6.58% 0.168 25.32 0.00000
> 2.6.0-test6-mm4 16 3.81 6.62% 0.193 83.97 0.00000
> 2.6.0-test7 16 3.89 6.57% 0.181 92.60 0.00000
>
> 2.6.0-test5 32 3.85 6.64% 0.159 3.38 0.00000
> 2.6.0-test5-mm2 32 3.80 6.75% 0.175 59.86 0.00000
> 2.6.0-test5-mm4 32 3.72 6.60% 0.180 55.04 0.00000
> 2.6.0-test6 32 3.80 6.73% 0.175 56.73 0.00000
> 2.6.0-test6-mm4 32 3.81 6.61% 0.171 55.65 0.00000
> 2.6.0-test7 32 3.60 5.97% 0.155 3.09 0.00000
>
> 2.6.0-test5 64 3.68 6.56% 0.161 3.90 0.00000
> 2.6.0-test5-mm2 64 3.60 6.51% 0.163 6.67 0.00000
> 2.6.0-test5-mm4 64 3.71 6.58% 0.166 5.44 0.00000
> 2.6.0-test6 64 3.53 6.32% 0.160 4.14 0.00000
> 2.6.0-test6-mm4 64 3.54 6.21% 0.185 106.22 0.00000
> 2.6.0-test7 64 3.61 6.15% 0.156 4.34 0.00000
>
> 2.6.0-test5 128 3.32 6.23% 0.166 4.42 0.00000
> 2.6.0-test5-mm2 128 3.64 6.70% 0.163 2.98 0.00000
> 2.6.0-test5-mm4 128 3.71 6.85% 0.169 5.37 0.00000
> 2.6.0-test6 128 3.44 6.25% 0.174 63.35 0.00000
> 2.6.0-test6-mm4 128 3.49 6.36% 0.174 64.77 0.00000
> 2.6.0-test7 128 3.43 6.15% 0.159 4.05 0.00000
>
> 2.6.0-test5 256 3.35 6.56% 0.161 3.65 0.00000
> 2.6.0-test5-mm2 256 3.29 6.41% 0.164 3.68 0.00000
> 2.6.0-test5-mm4 256 3.27 6.35% 0.166 4.84 0.00000
> 2.6.0-test6 256 2.97 5.80% 0.171 47.92 0.00000
> 2.6.0-test6-mm4 256 3.06 5.93% 0.159 3.02 0.00000
> 2.6.0-test7 256 3.11 5.88% 0.159 3.91 0.00000
>
>
> dbench did not regress.
>
> dbench-1.3 reiserfs 192 processes Average +/- MB/sec
> 2.6.0-test7 77.3 7.1
> 2.6.0-test6-mm4 76.6 7.2
> 2.6.0-test5-mm2 76.1 6.8
> 2.6.0-test6 73.0 5.1
> 2.6.0-test5-mm4 72.5 6.6
> 2.6.0-test5 65.0 1.2
>
> dbench-1.3 reiserfs 64 processes Average +/- MB/sec
> 2.6.0-test5-mm4 111.4 9.1
> 2.6.0-test6-mm4 109.6 7.5
> 2.6.0-test6 108.8 4.2
> 2.6.0-test5-mm2 104.4 8.1
> 2.6.0-test7 102.3 8.1
> 2.6.0-test5 100.6 4.7
>
> dbench-1.3 xfs 192 processes Average +/- MB/sec
> 2.6.0-test5 112.5 4.0
> 2.6.0-test5-mm2 110.8 1.6
> 2.6.0-test6-mm4 110.7 3.4
> 2.6.0-test5-mm4 109.7 2.0
> 2.6.0-test7 109.0 5.1
> 2.6.0-test6 108.5 6.1
>
> dbench-1.3 xfs 64 processes Average +/- MB/sec
> 2.6.0-test5 143.7 0.9
> 2.6.0-test7 143.6 2.1
> 2.6.0-test5-mm2 141.5 1.6
> 2.6.0-test6 141.4 0.5
> 2.6.0-test6-mm4 141.2 1.5
> 2.6.0-test5-mm4 140.0 1.4
>
> dbench-1.3 ext2 192 processes Average +/- MB/sec
> 2.6.0-test6 209.5 11.8
> 2.6.0-test7 205.2 20.8
> 2.6.0-test5-mm4 200.8 6.2
> 2.6.0-test5-mm2 200.0 10.3
> 2.6.0-test5 198.1 19.5
> 2.6.0-test6-mm4 193.4 12.0
>
> dbench-1.3 ext2 64 processes Average +/- MB/sec
> 2.6.0-test7 256.8 0.6
> 2.6.0-test6 256.7 1.6
> 2.6.0-test5 256.2 1.6
> 2.6.0-test5-mm4 255.5 1.4
> 2.6.0-test5-mm2 255.3 0.9
> 2.6.0-test6-mm4 255.3 1.7
>
> dbench-1.3 jfs 192 processes Average +/- MB/sec
> 2.6.0-test5-mm4 11.3 0.5
> 2.6.0-test7 11.1 0.2
> 2.6.0-test6-mm4 11.0 0.3
>
> dbench-1.3 jfs 64 processes Average +/- MB/sec
> 2.6.0-test5-mm2 20.0 2.9
> 2.6.0-test5 19.5 2.4
> 2.6.0-test5-mm4 18.6 2.4
> 2.6.0-test7 18.5 1.1
> 2.6.0-test6-mm4 18.1 1.6
>
> dbench-1.3 ext3 192 processes Average +/- MB/sec
> 2.6.0-test7 110.8 23.1
> 2.6.0-test5-mm2 107.5 39.7
> 2.6.0-test6 103.9 39.0
> 2.6.0-test6-mm4 99.7 38.7
> 2.6.0-test5-mm4 82.6 35.5
> 2.6.0-test5 73.6 15.1
>
> dbench-1.3 ext3 64 processes Average +/- MB/sec
> 2.6.0-test7 177.9 3.4
> 2.6.0-test6 175.1 3.6
> 2.6.0-test5 174.5 4.2
> 2.6.0-test6-mm4 174.2 2.5
> 2.6.0-test5-mm2 172.8 4.0
> 2.6.0-test5-mm4 172.6 2.3
>
>
> --
> Randy Hron
> http://home.earthlink.net/~rwhron/kernel/bigbox.html
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [BENCHMARK] I/O regression after 2.6.0-test5
2003-10-20 0:37 rwhron
@ 2003-10-20 1:35 ` Nick Piggin
2003-10-20 4:52 ` Andrew Morton
2003-10-20 4:51 ` Dave Olien
2003-10-24 13:52 ` Magnus Andersson
2 siblings, 1 reply; 25+ messages in thread
From: Nick Piggin @ 2003-10-20 1:35 UTC (permalink / raw)
To: rwhron; +Cc: linux-kernel, akpm
rwhron@earthlink.net wrote:
>There was about a 50% regression in jobs/minute in AIM7
>database workload on quad P3 Xeon. The CPU time has not
>gone up, so the extra run time is coming from something
>else. (I/O or I/O scheduler?)
>
>tiobench sequential reads has a significant regression too.
>
>Regression appears unrelated to filesystem type.
>
>dbench was not affected.
>
>The AIM7 was run on ext2.
>
Yeah I'd say its all due to the IO scheduler. There is a problem
I'm thinking about how to fix - its the likely cause of this too.
^ permalink raw reply [flat|nested] 25+ messages in thread
* [BENCHMARK] I/O regression after 2.6.0-test5
@ 2003-10-20 0:37 rwhron
2003-10-20 1:35 ` Nick Piggin
` (2 more replies)
0 siblings, 3 replies; 25+ messages in thread
From: rwhron @ 2003-10-20 0:37 UTC (permalink / raw)
To: linux-kernel; +Cc: akpm
There was about a 50% regression in jobs/minute in AIM7
database workload on quad P3 Xeon. The CPU time has not
gone up, so the extra run time is coming from something
else. (I/O or I/O scheduler?)
tiobench sequential reads has a significant regression too.
Regression appears unrelated to filesystem type.
dbench was not affected.
The AIM7 was run on ext2.
AIM7 dbase workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5-mm2 32 609.3 311.9 139.5
2.6.0-test5-mm4 32 595.2 319.4 131.7
2.6.0-test6-mm4 32 590.8 321.8 130.7
2.6.0-test5 32 590.2 322.0 131.9
2.6.0-test6 32 320.6 592.9 129.1
2.6.0-test7 32 290.3 654.8 120.9
2.6.0-test5-mm2 64 786.2 483.5 253.2
2.6.0-test6-mm4 64 779.4 487.8 253.8
2.6.0-test5 64 774.3 490.9 256.6
2.6.0-test5-mm4 64 772.0 492.4 253.7
2.6.0-test6 64 430.5 883.1 247.6
2.6.0-test7 64 402.6 944.2 231.5
2.6.0-test5-mm2 96 869.2 656.0 382.2
2.6.0-test6-mm4 96 867.6 657.2 387.4
2.6.0-test5-mm4 96 863.1 660.7 385.2
2.6.0-test5 96 859.0 663.9 389.5
2.6.0-test6 96 478.5 1191.7 368.1
2.6.0-test7 96 453.7 1256.9 346.8
2.6.0-test5-mm2 128 929.4 818.0 526.1
2.6.0-test6-mm4 128 922.6 824.1 523.3
2.6.0-test5-mm4 128 916.6 829.5 509.1
2.6.0-test5 128 910.3 835.2 528.6
2.6.0-test7 128 500.5 1519.1 461.3
2.6.0-test6 128 500.2 1520.1 497.4
2.6.0-test5-mm2 160 962.5 987.4 674.6
2.6.0-test5-mm4 160 952.7 997.6 655.5
2.6.0-test6-mm4 160 950.0 1000.5 659.5
2.6.0-test5 160 741.5 1281.7 644.8
2.6.0-test7 160 518.5 1833.1 589.6
2.6.0-test6 160 479.6 1981.8 618.9
2.6.0-test5-mm2 192 993.9 1147.5 810.0
2.6.0-test5-mm4 192 982.7 1160.5 788.2
2.6.0-test6-mm4 192 978.8 1165.2 808.4
2.6.0-test5 192 840.0 1357.8 776.0
2.6.0-test7 192 537.9 2120.2 709.3
2.6.0-test6 192 494.3 2307.1 748.2
2.6.0-test5-mm2 224 1011.2 1315.8 943.5
2.6.0-test5-mm4 224 1004.7 1324.3 942.1
2.6.0-test6-mm4 224 999.5 1331.2 920.0
2.6.0-test5 224 986.9 1348.2 911.8
2.6.0-test7 224 546.7 2433.8 832.8
2.6.0-test6 224 482.9 2755.1 862.1
2.6.0-test5-mm2 256 1030.9 1475.1 1076.5
2.6.0-test5-mm4 256 1027.1 1480.6 1071.4
2.6.0-test6-mm4 256 1010.3 1505.2 1074.8
2.6.0-test5 256 999.5 1521.3 1029.3
2.6.0-test7 256 556.6 2731.9 951.0
2.6.0-test6 256 545.1 2789.5 978.4
Similarly, AIM7 fileserver benchmark had a drop in throughput.
CPU time is down, but total runtime is way up.
AIM7 fserver workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5 4 79.7 304.0 38.3
2.6.0-test5-mm2 4 77.4 313.4 38.3
2.6.0-test5-mm4 4 76.5 317.1 36.7
2.6.0-test6-mm4 4 65.5 369.9 34.5
2.6.0-test7 4 45.5 532.5 32.2
2.6.0-test6 4 45.0 538.7 31.0
2.6.0-test5 8 129.8 373.5 66.3
2.6.0-test5-mm2 8 103.1 470.4 62.0
2.6.0-test5-mm4 8 102.2 474.2 61.8
2.6.0-test6 8 82.7 585.9 84.5
2.6.0-test6-mm4 8 82.4 588.1 61.1
2.6.0-test7 8 61.0 795.0 59.9
2.6.0-test5 12 164.1 443.0 93.1
2.6.0-test5-mm2 12 112.4 646.8 98.4
2.6.0-test5-mm4 12 111.2 653.8 96.7
2.6.0-test6-mm4 12 95.2 763.6 96.9
2.6.0-test6 12 87.9 827.6 118.9
2.6.0-test7 12 54.5 1335.0 91.8
2.6.0-test5 16 173.4 559.0 116.5
2.6.0-test5-mm2 16 128.8 752.9 119.9
2.6.0-test5-mm4 16 125.4 773.1 124.3
2.6.0-test6-mm4 16 104.2 930.3 118.7
2.6.0-test6 16 101.4 956.0 107.4
2.6.0-test7 16 70.6 1372.6 112.2
2.6.0-test5 20 180.5 671.5 143.0
2.6.0-test5-mm4 20 144.9 836.4 143.0
2.6.0-test5-mm2 20 141.7 855.3 144.1
2.6.0-test6 20 117.7 1030.0 131.1
2.6.0-test6-mm4 20 114.4 1059.5 143.0
2.6.0-test7 20 78.7 1539.9 134.0
2.6.0-test5 24 189.1 769.0 176.0
2.6.0-test5-mm2 24 156.2 930.9 180.2
2.6.0-test5-mm4 24 154.5 941.6 176.9
2.6.0-test6 24 134.5 1081.1 156.5
2.6.0-test6-mm4 24 124.5 1167.8 170.2
2.6.0-test7 24 76.9 1892.1 169.0
2.6.0-test5 28 204.6 829.2 197.5
2.6.0-test5-mm2 28 158.7 1069.2 209.8
2.6.0-test5-mm4 28 156.0 1087.7 211.5
2.6.0-test6-mm4 28 134.2 1264.0 196.9
2.6.0-test6 28 133.7 1268.7 174.7
2.6.0-test7 28 84.1 2016.4 188.5
2.6.0-test5 32 213.7 907.4 228.9
2.6.0-test5-mm2 32 151.9 1276.7 239.7
2.6.0-test5-mm4 32 151.5 1279.6 241.4
2.6.0-test6 32 147.0 1318.9 197.0
2.6.0-test6-mm4 32 138.8 1397.6 225.4
2.6.0-test7 32 88.5 2190.3 214.9
Same story for the shared workload. Jobs/minute is down a
lot since 2.6.0-test5.
AIM7 shared workload
kernel Tasks Jobs/Min Real CPU
2.6.0-test5 64 1888.8 197.2 168.3
2.6.0-test5-mm2 64 1517.6 245.4 177.0
2.6.0-test5-mm4 64 1488.6 250.2 180.5
2.6.0-test6-mm4 64 1239.0 300.6 175.0
2.6.0-test7 64 1225.1 304.0 166.6
2.6.0-test6 64 880.0 423.3 168.1
2.6.0-test5 128 2094.9 355.6 338.1
2.6.0-test5-mm2 128 1837.4 405.4 352.4
2.6.0-test5-mm4 128 1826.6 407.8 358.4
2.6.0-test6-mm4 128 1561.8 477.0 346.6
2.6.0-test7 128 1176.5 633.2 332.6
2.6.0-test6 128 1142.1 652.3 332.6
2.6.0-test5 192 2256.0 495.3 507.8
2.6.0-test5-mm2 192 2074.9 538.5 530.9
2.6.0-test5-mm4 192 2044.8 546.5 536.9
2.6.0-test6-mm4 192 1711.8 652.8 527.3
2.6.0-test6 192 1328.3 841.3 505.6
2.6.0-test7 192 1317.9 847.9 499.1
2.6.0-test5 256 2389.0 623.7 683.8
2.6.0-test5-mm2 256 2207.2 675.0 720.6
2.6.0-test5-mm4 256 2186.2 681.5 729.9
2.6.0-test6-mm4 256 1780.7 836.7 719.0
2.6.0-test6 256 1479.5 1007.0 672.9
2.6.0-test7 256 1282.0 1162.2 672.8
2.6.0-test5 320 2457.9 757.7 856.6
2.6.0-test5-mm4 320 2330.2 799.2 920.0
2.6.0-test5-mm2 320 2327.9 800.0 919.4
2.6.0-test6-mm4 320 1864.7 998.8 900.9
2.6.0-test6 320 1575.9 1181.8 853.5
2.6.0-test7 320 1317.9 1413.1 844.4
2.6.0-test5 384 2480.4 901.0 1036.5
2.6.0-test5-mm2 384 2440.2 915.9 1111.6
2.6.0-test5-mm4 384 2434.2 918.1 1111.0
2.6.0-test6-mm4 384 1883.9 1186.3 1091.7
2.6.0-test6 384 1629.8 1371.3 1026.3
2.6.0-test7 384 1326.3 1685.0 1018.6
2.6.0-test5 448 2529.8 1030.7 1209.2
2.6.0-test5-mm4 448 2511.6 1038.1 1306.8
2.6.0-test5-mm2 448 2510.0 1038.8 1306.1
2.6.0-test6-mm4 448 1934.3 1347.9 1283.6
2.6.0-test6 448 1612.3 1617.1 1203.3
2.6.0-test7 448 1344.7 1939.0 1200.9
2.6.0-test5-mm2 512 2579.1 1155.4 1504.0
2.6.0-test5 512 2566.2 1161.2 1386.2
2.6.0-test5-mm4 512 2552.9 1167.2 1509.2
2.6.0-test6-mm4 512 1921.5 1550.8 1475.1
2.6.0-test6 512 1538.6 1936.7 1376.9
2.6.0-test7 512 1370.5 2174.3 1379.2
tiobench read performance in test[67] has a big regression in throughput
as the number of threads goes up. The max and long latency numbers get
bad too. It appears to affect ext2, ext3, reiserfs, jfs and xfs.
Regression is in 2.6.0-test[67] and test6-mm4.
tiobench-0.3.3
Unit information
================
File size = 8192 megabytes
Blk Size = 4096 bytes
Rate = megabytes per second
CPU% = percentage of CPU used during the test
Latency = milliseconds
Lat% = percent of requests that took longer than X seconds
CPU Eff = Rate divided by CPU% - throughput per cpu load
Sequential Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 28.55 13.89% 0.408 754.59 0.00000
2.6.0-test5-mm2 1 28.66 13.98% 0.406 742.76 0.00000
2.6.0-test5-mm4 1 28.46 13.90% 0.409 887.02 0.00000
2.6.0-test6 1 28.47 13.81% 0.409 1069.74 0.00000
2.6.0-test6-mm4 1 28.41 13.89% 0.410 723.14 0.00000
2.6.0-test7 1 28.71 13.92% 0.405 715.02 0.00000
With one thread, performance is similar.
2.6.0-test5 8 34.01 17.09% 2.620 834.85 0.00000
2.6.0-test5-mm2 8 32.44 16.44% 2.765 844.99 0.00000
2.6.0-test5-mm4 8 30.75 15.55% 2.967 952.18 0.00000
2.6.0-test6 8 20.67 9.90% 4.477 1569.30 0.00000
2.6.0-test6-mm4 8 21.55 10.31% 4.260 1790.19 0.00000
2.6.0-test7 8 21.46 10.24% 4.312 1913.24 0.00000
2.6.0-test5 16 30.13 15.04% 5.980 1337.69 0.00000
2.6.0-test5-mm2 16 29.71 14.97% 6.024 1115.95 0.00000
2.6.0-test5-mm4 16 27.10 13.65% 6.606 1008.67 0.00000
2.6.0-test6 16 16.44 7.90% 11.201 2820.85 0.00000
2.6.0-test6-mm4 16 16.14 7.78% 11.382 2753.13 0.00000
2.6.0-test7 16 18.34 8.64% 9.877 2742.49 0.00000
2.6.0-test5 32 29.46 14.62% 11.954 1560.02 0.00000
2.6.0-test5-mm2 32 28.01 14.26% 12.808 1627.00 0.00000
2.6.0-test5-mm4 32 28.78 14.63% 12.182 1859.56 0.00000
2.6.0-test6 32 15.60 7.55% 23.415 4877.75 0.00000
2.6.0-test6-mm4 32 14.63 7.20% 24.713 4846.63 0.00000
2.6.0-test7 32 14.79 7.14% 24.234 4855.64 0.00000
2.6.0-test5 64 25.24 13.57% 26.901 1743.05 0.00000
2.6.0-test5-mm2 64 24.01 13.27% 27.922 1849.82 0.00000
2.6.0-test5-mm4 64 23.67 13.15% 28.513 1956.58 0.00000
2.6.0-test6 64 9.45 5.24% 75.491 9578.85 2.75073
2.6.0-test6-mm4 64 8.51 4.82% 84.118 13166.74 3.11647
2.6.0-test7 64 8.58 4.86% 83.430 13622.86 3.05581
2.6.0-test5 128 10.41 9.76% 129.662 1820.78 0.00000
2.6.0-test5-mm2 128 11.00 10.44% 120.118 1742.44 0.00000
2.6.0-test5-mm4 128 10.67 10.15% 122.871 1912.79 0.00000
2.6.0-test6 128 2.97 2.66% 472.251 25694.03 8.39324
2.6.0-test6-mm4 128 2.94 2.66% 473.943 25530.14 8.39028
2.6.0-test7 128 2.94 2.68% 480.175 26306.38 8.39257
At 256 threads, throughput is down and max latency skyrockets.
2.6.0-test5 256 9.13 13.30% 305.097 1446.43 0.00000
2.6.0-test5-mm2 256 9.09 13.60% 304.019 1648.80 0.00000
2.6.0-test5-mm4 256 8.63 13.66% 325.793 1542.62 0.00000
2.6.0-test6 256 2.71 2.49% 937.603 529116.60 8.94007
2.6.0-test6-mm4 256 2.62 2.46% 979.221 785760.77 9.12519
2.6.0-test7 256 2.56 2.39% 988.719 582067.86 9.28287
Random Reads ext2
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 0.95 0.94% 12.373 114.89 0.00000
2.6.0-test5-mm2 1 0.90 0.88% 13.061 113.19 0.00000
2.6.0-test5-mm4 1 0.53 0.60% 21.903 151.35 0.00000
2.6.0-test6 1 0.73 0.82% 16.053 391.29 0.00000
2.6.0-test6-mm4 1 0.76 0.81% 15.388 126.51 0.00000
2.6.0-test7 1 0.96 0.94% 12.192 101.93 0.00000
2.6.0-test5 8 4.11 4.58% 19.433 137.74 0.00000
2.6.0-test5-mm2 8 3.94 4.03% 20.931 144.16 0.00000
2.6.0-test5-mm4 8 2.81 3.28% 30.086 222.27 0.00000
2.6.0-test6 8 3.25 3.21% 27.598 640.41 0.00000
2.6.0-test6-mm4 8 3.98 3.82% 22.078 324.87 0.00000
2.6.0-test7 8 3.33 3.27% 27.069 417.17 0.00000
2.6.0-test5 16 4.05 3.90% 42.424 236.49 0.00000
2.6.0-test5-mm2 16 4.09 4.31% 41.651 255.17 0.00000
2.6.0-test5-mm4 16 3.05 3.19% 55.847 332.10 0.00000
2.6.0-test6 16 3.43 3.28% 48.836 934.02 0.00000
2.6.0-test6-mm4 16 3.21 2.69% 52.841 936.50 0.00000
2.6.0-test7 16 4.93 4.17% 34.720 630.63 0.00000
2.6.0-test5 32 4.41 3.58% 70.910 516.28 0.00000
2.6.0-test5-mm2 32 4.23 4.17% 78.492 372.63 0.00000
2.6.0-test5-mm4 32 4.19 3.90% 76.001 472.90 0.00000
2.6.0-test6 32 3.74 3.20% 92.407 1578.33 0.00000
2.6.0-test6-mm4 32 2.94 2.97% 116.663 1722.23 0.00000
2.6.0-test7 32 2.62 2.36% 129.093 2248.32 0.00000
2.6.0-test5 64 4.47 4.00% 137.277 1077.53 0.00000
2.6.0-test5-mm2 64 4.79 4.53% 124.035 1144.38 0.00000
2.6.0-test5-mm4 64 4.13 3.81% 143.321 1265.44 0.00000
2.6.0-test6 64 3.89 3.62% 166.555 2679.86 0.00000
2.6.0-test6-mm4 64 3.23 3.15% 200.698 3160.57 0.00000
2.6.0-test7 64 3.14 2.74% 206.849 3384.30 0.00000
2.6.0-test5 128 4.86 4.83% 243.704 2287.09 0.00000
2.6.0-test5-mm2 128 5.08 4.85% 228.102 2391.21 0.00000
2.6.0-test5-mm4 128 4.16 4.65% 270.840 2501.04 0.00000
2.6.0-test6 128 3.19 3.09% 389.583 6092.04 0.12601
2.6.0-test6-mm4 128 2.75 2.78% 439.379 6632.53 0.65524
2.6.0-test7 128 3.16 3.16% 388.762 5810.95 0.07560
2.6.0-test5 256 5.51 7.00% 427.503 2511.79 0.00000
2.6.0-test5-mm2 256 5.23 7.39% 449.007 3169.28 0.00000
2.6.0-test5-mm4 256 4.48 6.76% 526.994 3618.17 0.00000
2.6.0-test6 256 2.01 2.44% 1156.505 14117.02 16.82291
2.6.0-test6-mm4 256 1.99 2.38% 1169.942 13903.51 16.56250
2.6.0-test7 256 2.09 2.41% 1133.624 13195.27 14.97396
Sequential and random writes on ext2 are not regressed.
ext3 has a regression similar to ext2 for sequential and random reads.
Sequential Reads ext3
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 29.03 14.00% 0.401 745.26 0.00000
2.6.0-test5-mm2 1 28.05 13.82% 0.415 763.49 0.00000
2.6.0-test5-mm4 1 29.31 14.10% 0.397 867.89 0.00000
2.6.0-test6 1 28.92 13.83% 0.402 747.59 0.00000
2.6.0-test6-mm4 1 28.01 13.67% 0.415 1162.40 0.00000
2.6.0-test7 1 27.95 13.54% 0.416 860.38 0.00000
2.6.0-test5 8 14.53 8.42% 6.187 1280.16 0.00000
2.6.0-test5-mm2 8 14.25 8.56% 6.428 1438.12 0.00000
2.6.0-test5-mm4 8 14.42 8.60% 6.317 1335.28 0.00000
2.6.0-test6 8 10.37 6.00% 8.791 2065.27 0.00000
2.6.0-test6-mm4 8 9.38 5.81% 9.765 1978.67 0.00000
2.6.0-test7 8 10.25 6.02% 8.931 1977.71 0.00000
2.6.0-test5 16 11.38 7.09% 15.604 1387.68 0.00000
2.6.0-test5-mm2 16 12.84 7.97% 14.206 1246.21 0.00000
2.6.0-test5-mm4 16 12.89 8.00% 14.050 1790.63 0.00000
2.6.0-test6 16 5.13 3.45% 34.925 3982.64 0.00000
2.6.0-test6-mm4 16 4.91 3.42% 37.057 4349.95 0.00000
2.6.0-test7 16 5.27 3.56% 34.774 3852.71 0.00000
2.6.0-test5 32 8.77 6.17% 41.461 1856.08 0.00000
2.6.0-test5-mm2 32 8.79 6.26% 40.822 1667.85 0.00000
2.6.0-test5-mm4 32 8.74 6.41% 41.519 1797.28 0.00000
2.6.0-test6 32 3.32 2.60% 107.252 9267.00 0.00649
2.6.0-test6-mm4 32 3.58 2.78% 100.364 8453.15 0.00496
2.6.0-test7 32 3.39 2.65% 104.283 9029.86 0.00367
2.6.0-test5 64 6.29 5.30% 113.016 1947.14 0.00000
2.6.0-test5-mm2 64 7.58 6.00% 81.717 1722.83 0.00000
2.6.0-test5-mm4 64 7.63 6.07% 79.882 1709.56 0.00000
2.6.0-test6 64 2.63 2.25% 265.280 18736.77 8.76222
2.6.0-test6-mm4 64 2.43 2.17% 277.430 17986.61 8.95066
2.6.0-test7 64 2.40 2.16% 286.947 17890.59 9.59925
2.6.0-test5 128 6.19 6.61% 216.851 1852.30 0.00000
2.6.0-test5-mm2 128 6.73 7.19% 191.761 2049.82 0.00000
2.6.0-test5-mm4 128 6.09 6.55% 209.661 1910.36 0.00000
2.6.0-test6 128 2.06 1.95% 640.506 32660.27 12.41284
2.6.0-test6-mm4 128 2.03 1.98% 645.423 30932.33 12.57601
2.6.0-test7 128 2.05 2.02% 639.791 30885.16 12.72559
2.6.0-test5 256 6.15 9.88% 457.029 1880.47 0.00000
2.6.0-test5-mm2 256 5.51 9.06% 518.943 2002.06 0.00000
2.6.0-test5-mm4 256 5.30 8.81% 535.291 2103.11 0.00000
2.6.0-test6 256 1.85 1.78% 1293.732 230380.55 13.73657
2.6.0-test6-mm4 256 1.91 1.90% 1246.640 243088.10 13.26160
2.6.0-test7 256 1.81 1.78% 1351.071 281394.59 13.80543
Random Reads ext3
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 0.90 0.85% 12.963 112.10 0.00000
2.6.0-test5-mm2 1 0.93 0.97% 12.637 125.35 0.00000
2.6.0-test5-mm4 1 0.85 0.95% 13.714 122.66 0.00000
2.6.0-test6 1 0.73 0.71% 16.054 516.87 0.00000
2.6.0-test6-mm4 1 0.91 0.91% 12.907 125.50 0.00000
2.6.0-test7 1 0.94 0.82% 12.443 110.45 0.00000
2.6.0-test5 8 3.78 4.24% 21.477 131.88 0.00000
2.6.0-test5-mm2 8 4.01 4.49% 20.218 124.38 0.00000
2.6.0-test5-mm4 8 4.43 4.63% 18.684 142.49 0.00000
2.6.0-test6 8 3.55 3.11% 23.596 362.51 0.00000
2.6.0-test6-mm4 8 2.52 2.69% 32.927 534.18 0.00000
2.6.0-test7 8 4.21 3.80% 19.881 459.00 0.00000
2.6.0-test5 16 4.10 3.74% 37.249 236.48 0.00000
2.6.0-test5-mm2 16 4.42 4.48% 37.848 208.96 0.00000
2.6.0-test5-mm4 16 4.25 4.51% 36.874 220.75 0.00000
2.6.0-test6 16 3.27 3.13% 49.687 803.39 0.00000
2.6.0-test6-mm4 16 3.37 3.17% 49.112 951.18 0.00000
2.6.0-test7 16 4.49 3.87% 38.468 727.87 0.00000
2.6.0-test5 32 4.28 4.06% 76.377 458.79 0.00000
2.6.0-test5-mm2 32 4.53 4.40% 71.762 466.83 0.00000
2.6.0-test5-mm4 32 4.48 4.23% 70.711 361.46 0.00000
2.6.0-test6 32 3.35 3.06% 98.621 1728.18 0.00000
2.6.0-test6-mm4 32 4.13 3.54% 77.004 1321.89 0.00000
2.6.0-test7 32 3.95 3.30% 79.804 1402.47 0.00000
2.6.0-test5 64 4.89 4.80% 134.964 1243.33 0.00000
2.6.0-test5-mm2 64 4.63 4.17% 108.944 1006.23 0.00000
2.6.0-test5-mm4 64 3.70 3.92% 132.534 1222.82 0.00000
2.6.0-test6 64 3.35 2.98% 189.642 2847.96 0.00000
2.6.0-test6-mm4 64 3.12 3.03% 207.658 3362.27 0.00000
2.6.0-test7 64 3.02 2.89% 212.838 2652.28 0.00000
2.6.0-test5 128 5.26 5.11% 230.823 2088.49 0.00000
2.6.0-test5-mm2 128 4.77 4.65% 252.354 2283.06 0.00000
2.6.0-test5-mm4 128 4.07 4.26% 280.692 2923.29 0.00000
2.6.0-test6 128 2.88 3.07% 431.937 5967.29 0.12601
2.6.0-test6-mm4 128 3.09 3.26% 403.965 5559.31 0.02520
2.6.0-test7 128 3.01 2.95% 407.900 6243.88 0.10081
2.6.0-test5 256 5.18 6.62% 463.528 3270.79 0.00000
2.6.0-test5-mm2 256 5.20 7.12% 452.987 3055.03 0.00000
2.6.0-test5-mm4 256 4.36 5.89% 539.848 3465.46 0.00000
2.6.0-test6 256 2.26 2.92% 1013.076 11495.98 12.47396
2.6.0-test6-mm4 256 2.49 2.98% 943.926 10902.18 10.72917
2.6.0-test7 256 2.43 3.04% 996.420 11016.00 12.08334
Sequential Writes ext3
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 56.15 64.55% 0.170 5700.14 0.00010
2.6.0-test5-mm2 1 56.14 67.38% 0.173 4383.06 0.00005
2.6.0-test5-mm4 1 55.64 66.40% 0.173 4323.48 0.00000
2.6.0-test6 1 56.15 65.22% 0.171 4578.10 0.00005
2.6.0-test6-mm4 1 55.81 66.36% 0.172 5623.30 0.00010
2.6.0-test7 1 55.94 65.10% 0.171 4544.88 0.00000
2.6.0-test5 8 2.99 4.94% 23.859 1119399.58 0.08983
2.6.0-test5-mm2 8 3.00 5.08% 24.803 613704.08 0.09155
2.6.0-test5-mm4 8 2.98 5.06% 25.599 515385.61 0.09627
2.6.0-test6 8 4.65 7.53% 15.776 347393.72 0.07853
2.6.0-test6-mm4 8 3.02 5.04% 25.421 399480.58 0.09512
2.6.0-test7 8 3.12 5.12% 23.991 197721.86 0.09255
2.6.0-test5 16 2.58 4.55% 55.284 715963.30 0.14072
2.6.0-test5-mm2 16 2.93 5.09% 50.697 648332.45 0.11883
2.6.0-test5-mm4 16 3.21 5.62% 47.072 544422.32 0.12240
2.6.0-test6 16 3.45 6.12% 39.175 424983.10 0.11596
2.6.0-test6-mm4 16 2.72 4.76% 54.117 429887.30 0.13976
2.6.0-test7 16 2.74 4.73% 52.594 886672.08 0.13404
2.6.0-test5 32 2.28 4.43% 127.746 833586.10 0.20012
2.6.0-test5-mm2 32 2.25 4.58% 124.219 1054925.33 0.20647
2.6.0-test5-mm4 32 2.11 4.33% 139.662 787258.98 0.22011
2.6.0-test6 32 2.80 5.35% 107.223 835055.83 0.18424
2.6.0-test6-mm4 32 2.28 4.26% 117.722 754407.32 0.18582
2.6.0-test7 32 2.21 4.38% 128.475 829790.72 0.21805
2.6.0-test5 64 2.02 4.54% 265.546 1322591.44 0.40007
2.6.0-test5-mm2 64 2.02 4.42% 278.978 1434418.75 0.42186
2.6.0-test5-mm4 64 2.25 4.83% 242.598 1592337.06 0.35062
2.6.0-test6 64 2.44 5.25% 222.436 1233112.15 0.38619
2.6.0-test6-mm4 64 1.98 4.45% 274.091 1442328.98 0.45623
2.6.0-test7 64 2.09 4.37% 249.893 1456365.96 0.30308
2.6.0-test5 128 2.21 4.96% 444.935 2758997.12 0.60168
2.6.0-test5-mm2 128 1.93 4.62% 568.801 3209338.67 0.84395
2.6.0-test5-mm4 128 1.99 4.67% 518.813 2778367.36 0.77595
2.6.0-test6 128 2.00 4.47% 497.636 2933240.88 0.65131
2.6.0-test6-mm4 128 2.03 4.71% 505.696 2839390.84 0.74044
2.6.0-test7 128 2.00 4.69% 491.950 2994053.88 0.75612
2.6.0-test5 256 2.11 7.77% 941.664 4873713.39 1.83378
2.6.0-test5-mm2 256 1.91 6.85% 1093.820 5897453.70 1.80015
2.6.0-test5-mm4 256 1.97 6.74% 1017.367 5695863.45 1.77222
2.6.0-test6 256 1.92 6.71% 1065.957 6104195.68 1.87802
2.6.0-test6-mm4 256 1.93 7.02% 1038.493 5678570.32 1.92114
2.6.0-test7 256 1.85 7.36% 1083.836 5720266.70 1.89524
Random Writes ext3
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 3.39 3.44% 0.399 37.37 0.00000
2.6.0-test5-mm2 1 3.22 3.38% 0.516 40.33 0.00000
2.6.0-test5-mm4 1 3.56 3.72% 0.081 10.61 0.00000
2.6.0-test6 1 2.34 2.39% 1.780 24.35 0.00000
2.6.0-test6-mm4 1 3.00 3.19% 0.746 37.62 0.00000
2.6.0-test7 1 3.05 3.06% 0.653 54.78 0.00000
2.6.0-test5 8 3.17 3.59% 13.525 1655.65 0.00000
2.6.0-test5-mm2 8 3.05 3.44% 14.349 2391.37 0.00000
2.6.0-test5-mm4 8 3.08 3.54% 14.093 1401.73 0.00000
2.6.0-test6 8 2.79 3.15% 14.906 3957.84 0.02500
2.6.0-test6-mm4 8 2.88 3.06% 15.029 1283.97 0.00000
2.6.0-test7 8 3.18 3.54% 14.407 1671.73 0.00000
2.6.0-test5 16 3.18 3.49% 26.252 1693.61 0.00000
2.6.0-test5-mm2 16 3.28 3.73% 25.573 3583.93 0.02500
2.6.0-test5-mm4 16 3.21 3.84% 25.857 6012.15 0.02500
2.6.0-test6 16 2.81 2.81% 30.448 2511.41 0.00000
2.6.0-test6-mm4 16 3.17 4.08% 20.815 5062.41 0.02500
2.6.0-test7 16 3.14 3.43% 28.987 2120.35 0.00000
2.6.0-test5 32 3.26 4.05% 46.822 3656.96 0.00000
2.6.0-test5-mm2 32 3.25 4.28% 47.088 2834.74 0.00000
2.6.0-test5-mm4 32 3.29 4.49% 44.765 2409.89 0.00000
2.6.0-test6 32 2.67 2.91% 62.319 11807.07 0.05000
2.6.0-test6-mm4 32 2.98 3.57% 57.025 2250.13 0.00000
2.6.0-test7 32 3.06 3.75% 50.252 9420.48 0.05000
2.6.0-test5 64 3.40 4.96% 46.634 2426.05 0.00000
2.6.0-test5-mm2 64 3.24 4.72% 93.293 7875.05 1.86492
2.6.0-test5-mm4 64 3.24 4.44% 79.998 5292.84 0.63004
2.6.0-test6 64 2.50 3.22% 137.484 6723.01 0.25202
2.6.0-test6-mm4 64 2.83 3.92% 79.495 9204.46 0.68044
2.6.0-test7 64 2.85 4.00% 91.255 6050.08 0.63004
2.6.0-test5 128 3.46 5.89% 101.597 6086.33 0.80645
2.6.0-test5-mm2 128 3.39 5.71% 129.468 6234.78 1.28528
2.6.0-test5-mm4 128 3.27 5.82% 128.846 6601.87 0.65524
2.6.0-test6 128 2.63 4.26% 123.641 6247.05 2.09173
2.6.0-test6-mm4 128 2.69 4.58% 100.787 4669.66 0.00000
2.6.0-test7 128 2.71 4.58% 91.826 4953.82 0.00000
2.6.0-test5 256 3.45 6.88% 210.999 10063.95 6.06770
2.6.0-test5-mm2 256 3.45 6.80% 263.620 8774.35 8.25520
2.6.0-test5-mm4 256 3.33 6.91% 209.373 10921.27 5.10416
2.6.0-test6 256 2.58 5.45% 353.922 7824.45 5.31250
2.6.0-test6-mm4 256 2.99 6.08% 152.068 6246.05 2.36979
2.6.0-test7 256 2.96 6.15% 69.124 3923.92 0.00000
Sequential Reads jfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 28.93 13.16% 0.402 958.45 0.00000
2.6.0-test5-mm2 1 28.84 13.33% 0.403 703.16 0.00000
2.6.0-test5-mm4 1 29.03 13.50% 0.401 871.93 0.00000
2.6.0-test6-mm4 1 28.90 13.54% 0.402 724.31 0.00000
2.6.0-test7 1 29.07 13.44% 0.400 677.09 0.00000
2.6.0-test5 8 50.81 24.01% 1.815 942.91 0.00000
2.6.0-test5-mm2 8 48.22 23.30% 1.918 959.38 0.00000
2.6.0-test5-mm4 8 46.01 21.88% 1.988 840.79 0.00000
2.6.0-test6-mm4 8 20.85 9.38% 4.417 2073.30 0.00000
2.6.0-test7 8 20.88 9.17% 4.402 1941.85 0.00000
2.6.0-test5 16 37.26 17.05% 5.004 855.09 0.00000
2.6.0-test5-mm2 16 40.27 18.83% 4.628 1024.08 0.00000
2.6.0-test5-mm4 16 38.68 18.07% 4.831 820.64 0.00000
2.6.0-test6-mm4 16 11.10 5.05% 16.392 3579.95 0.00000
2.6.0-test7 16 11.98 5.35% 15.258 3381.34 0.00000
2.6.0-test5 32 35.48 16.56% 10.439 1742.70 0.00000
2.6.0-test5-mm2 32 35.46 17.01% 10.292 1979.97 0.00000
2.6.0-test5-mm4 32 33.58 15.89% 11.012 1885.26 0.00000
2.6.0-test6-mm4 32 10.45 4.79% 34.665 6205.82 0.00743
2.6.0-test7 32 10.82 4.85% 33.369 6185.50 0.00668
2.6.0-test5 64 36.58 17.71% 19.855 3297.14 0.00000
2.6.0-test5-mm2 64 34.64 17.14% 21.114 3190.91 0.00000
2.6.0-test5-mm4 64 34.98 17.66% 20.673 3254.24 0.00000
2.6.0-test6-mm4 64 10.64 4.87% 67.501 11935.89 2.07191
2.6.0-test7 64 11.33 5.08% 61.240 11701.91 1.98450
2.6.0-test5 128 11.37 8.64% 109.981 2829.31 0.00000
2.6.0-test5-mm2 128 17.63 11.71% 64.633 2799.09 0.00000
2.6.0-test5-mm4 128 15.92 11.40% 73.376 3032.50 0.00000
2.6.0-test6-mm4 128 2.71 2.52% 520.381 33565.98 8.79664
2.6.0-test7 128 3.09 2.69% 409.208 33147.84 7.51347
2.6.0-test5 256 7.53 11.50% 348.180 2671.65 0.00000
2.6.0-test5-mm2 256 7.77 11.45% 319.292 2285.73 0.00000
2.6.0-test5-mm4 256 7.23 11.94% 364.323 2244.99 0.00000
2.6.0-test6-mm4 256 2.00 1.95% 1023.629 216889.13 12.88433
2.6.0-test7 256 1.85 1.88% 1090.929 455893.85 13.53616
Random Reads jfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 0.95 0.82% 12.368 86.79 0.00000
2.6.0-test5-mm2 1 0.97 0.89% 12.013 108.08 0.00000
2.6.0-test5-mm4 1 0.99 0.99% 11.796 110.46 0.00000
2.6.0-test6-mm4 1 0.98 0.91% 11.987 112.94 0.00000
2.6.0-test7 1 0.96 0.83% 12.187 100.56 0.00000
2.6.0-test5 8 4.68 4.23% 18.784 137.13 0.00000
2.6.0-test5-mm2 8 4.06 3.84% 21.942 145.02 0.00000
2.6.0-test5-mm4 8 4.01 3.80% 20.854 150.03 0.00000
2.6.0-test6-mm4 8 0.81 0.61% 112.779 1277.22 0.00000
2.6.0-test7 8 0.82 0.66% 110.357 1250.05 0.00000
2.6.0-test5 16 3.98 3.12% 44.566 188.99 0.00000
2.6.0-test5-mm2 16 4.35 3.42% 40.240 189.97 0.00000
2.6.0-test5-mm4 16 4.28 3.37% 41.846 199.23 0.00000
2.6.0-test6-mm4 16 0.75 0.62% 231.931 2631.93 0.00000
2.6.0-test7 16 0.84 0.65% 212.690 2578.59 0.00000
2.6.0-test5 32 4.09 2.87% 86.440 373.37 0.00000
2.6.0-test5-mm2 32 3.86 3.07% 91.487 443.48 0.00000
2.6.0-test5-mm4 32 3.88 3.07% 90.965 553.03 0.00000
2.6.0-test6-mm4 32 0.76 0.64% 455.886 5230.06 0.00000
2.6.0-test7 32 0.71 0.57% 477.112 5188.75 0.00000
2.6.0-test5 64 4.04 2.88% 169.495 1553.14 0.00000
2.6.0-test5-mm2 64 3.84 3.25% 170.017 1413.19 0.00000
2.6.0-test5-mm4 64 3.99 3.37% 167.645 1628.57 0.00000
2.6.0-test6-mm4 64 0.70 0.61% 950.464 10451.50 26.76412
2.6.0-test7 64 0.71 0.63% 923.148 10352.12 26.38609
2.6.0-test5 128 4.32 3.83% 268.303 2322.59 0.00000
2.6.0-test5-mm2 128 4.18 3.95% 253.072 2451.90 0.00000
2.6.0-test5-mm4 128 4.05 3.95% 255.242 2232.10 0.00000
2.6.0-test6-mm4 128 0.76 0.77% 1626.585 26333.82 35.78629
2.6.0-test7 128 0.78 0.69% 1521.032 24112.21 33.84576
2.6.0-test5 256 4.03 5.43% 585.871 3402.15 0.00000
2.6.0-test5-mm2 256 4.06 5.49% 556.512 3436.30 0.00000
2.6.0-test5-mm4 256 4.11 5.47% 543.649 2806.43 0.00000
2.6.0-test6-mm4 256 0.82 0.87% 2318.070 33523.86 44.84375
2.6.0-test7 256 0.81 0.85% 2315.106 34969.85 44.94791
Sequential Writes jfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 56.47 47.37% 0.170 3391.01 0.00000
2.6.0-test5-mm2 1 56.47 49.42% 0.172 2901.46 0.00000
2.6.0-test5-mm4 1 56.40 48.44% 0.181 2578.21 0.00000
2.6.0-test6-mm4 1 56.28 47.99% 0.175 2442.78 0.00000
2.6.0-test7 1 56.39 46.73% 0.171 3405.20 0.00000
2.6.0-test5 8 56.19 65.59% 1.343 8360.10 0.00100
2.6.0-test5-mm2 8 55.94 66.86% 1.336 17088.51 0.00162
2.6.0-test5-mm4 8 56.03 67.52% 1.316 12716.34 0.00105
2.6.0-test6-mm4 8 56.14 65.15% 1.320 13881.60 0.00100
2.6.0-test7 8 56.23 64.40% 1.335 11973.67 0.00109
2.6.0-test5 16 55.15 67.48% 2.651 30109.56 0.03486
2.6.0-test5-mm2 16 55.93 68.92% 2.576 39135.68 0.03081
2.6.0-test5-mm4 16 54.88 69.40% 2.679 36823.40 0.03596
2.6.0-test6-mm4 16 54.85 65.44% 2.641 21368.63 0.04229
2.6.0-test7 16 55.84 68.65% 2.635 63526.06 0.03019
2.6.0-test5 32 54.38 69.36% 5.204 96795.80 0.10614
2.6.0-test5-mm2 32 55.11 70.01% 4.989 55045.13 0.10557
2.6.0-test5-mm4 32 53.92 70.42% 5.099 68154.16 0.10362
2.6.0-test6-mm4 32 54.17 70.60% 5.074 81693.83 0.09809
2.6.0-test7 32 54.75 69.14% 4.998 108095.44 0.09742
2.6.0-test5 64 54.58 75.03% 9.331 208382.29 0.10910
2.6.0-test5-mm2 64 54.44 69.22% 9.195 156878.93 0.11329
2.6.0-test5-mm4 64 53.35 72.64% 9.621 158954.21 0.10639
2.6.0-test6-mm4 64 54.54 70.28% 9.386 148888.99 0.11191
2.6.0-test7 64 54.67 69.68% 8.969 189127.28 0.10605
2.6.0-test5 128 10.39 18.85% 93.764 2075500.73 0.09913
2.6.0-test5-mm2 128 15.87 27.47% 61.244 1306699.59 0.08931
2.6.0-test5-mm4 128 16.06 28.24% 56.744 1437340.48 0.08950
2.6.0-test6-mm4 128 20.89 35.44% 42.971 1018802.82 0.10410
2.6.0-test7 128 13.63 23.07% 69.588 1685679.31 0.08650
2.6.0-test5 256 10.65 20.80% 168.683 1900450.38 0.20290
2.6.0-test5-mm2 256 9.72 18.90% 202.289 2024501.68 0.13385
2.6.0-test5-mm4 256 8.52 17.57% 221.659 1967027.22 0.19260
2.6.0-test6-mm4 256 9.20 17.32% 200.496 1843020.23 0.12737
2.6.0-test7 256 8.85 17.47% 212.298 2080395.84 0.11768
Random Writes jfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 3.90 3.48% 0.060 0.56 0.00000
2.6.0-test5-mm2 1 3.89 3.25% 0.061 0.28 0.00000
2.6.0-test5-mm4 1 3.86 3.23% 0.061 0.35 0.00000
2.6.0-test6-mm4 1 3.93 3.40% 0.060 0.37 0.00000
2.6.0-test7 1 3.69 3.03% 0.059 0.88 0.00000
2.6.0-test5 8 3.86 6.54% 0.170 59.53 0.00000
2.6.0-test5-mm2 8 3.68 6.19% 0.170 77.58 0.00000
2.6.0-test5-mm4 8 3.69 6.03% 0.171 69.55 0.00000
2.6.0-test6-mm4 8 3.55 5.93% 0.153 34.88 0.00000
2.6.0-test7 8 3.56 5.99% 0.162 52.74 0.00000
2.6.0-test5 16 3.71 6.47% 0.162 43.67 0.00000
2.6.0-test5-mm2 16 3.82 6.63% 0.166 54.86 0.00000
2.6.0-test5-mm4 16 3.71 6.33% 0.168 43.72 0.00000
2.6.0-test6-mm4 16 3.42 6.01% 0.151 14.21 0.00000
2.6.0-test7 16 3.56 5.94% 0.149 4.41 0.00000
2.6.0-test5 32 3.71 6.48% 0.178 111.32 0.00000
2.6.0-test5-mm2 32 3.50 5.99% 0.148 3.16 0.00000
2.6.0-test5-mm4 32 3.56 6.47% 0.170 55.28 0.00000
2.6.0-test6-mm4 32 3.44 6.09% 0.147 4.23 0.00000
2.6.0-test7 32 3.22 5.41% 0.158 59.51 0.00000
2.6.0-test5 64 3.50 6.28% 0.148 3.96 0.00000
2.6.0-test5-mm2 64 3.34 5.99% 0.160 51.30 0.00000
2.6.0-test5-mm4 64 3.58 6.46% 0.157 7.48 0.00000
2.6.0-test6-mm4 64 3.25 5.67% 0.146 3.61 0.00000
2.6.0-test7 64 3.28 5.74% 0.168 55.67 0.00000
2.6.0-test5 128 3.28 5.72% 20.122 1415.53 0.00000
2.6.0-test5-mm2 128 3.29 5.92% 1.297 1229.30 0.00000
2.6.0-test5-mm4 128 3.42 6.33% 1.499 541.00 0.00000
2.6.0-test6-mm4 128 3.08 5.62% 1.341 1089.07 0.00000
2.6.0-test7 128 3.13 5.51% 0.595 923.55 0.00000
2.6.0-test5 256 2.80 5.18% 89.100 3075.82 0.00000
2.6.0-test5-mm2 256 2.98 5.76% 33.288 2094.84 0.00000
2.6.0-test5-mm4 256 3.19 6.47% 13.738 1648.72 0.00000
2.6.0-test6-mm4 256 2.65 5.49% 18.498 2632.76 0.00000
2.6.0-test7 256 2.74 5.45% 10.049 2382.47 0.00000
Sequential Reads reiserfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 28.20 15.37% 0.413 97.43 0.00000
2.6.0-test5-mm2 1 30.21 16.13% 0.385 1159.97 0.00000
2.6.0-test5-mm4 1 28.15 15.29% 0.413 771.52 0.00000
2.6.0-test6 1 28.84 15.63% 0.404 111.79 0.00000
2.6.0-test6-mm4 1 29.99 16.01% 0.388 1065.08 0.00000
2.6.0-test7 1 29.45 15.70% 0.395 113.20 0.00000
2.6.0-test5 8 29.77 17.07% 2.974 854.50 0.00000
2.6.0-test5-mm2 8 31.14 18.03% 2.811 653.61 0.00000
2.6.0-test5-mm4 8 32.23 18.70% 2.727 542.20 0.00000
2.6.0-test6 8 21.41 11.92% 4.275 1487.18 0.00000
2.6.0-test6-mm4 8 21.93 12.30% 4.185 1320.72 0.00000
2.6.0-test7 8 21.40 11.78% 4.263 1261.39 0.00000
2.6.0-test5 16 29.99 17.01% 5.829 901.20 0.00000
2.6.0-test5-mm2 16 29.66 17.05% 5.954 1288.12 0.00000
2.6.0-test5-mm4 16 30.20 17.66% 5.872 1153.05 0.00000
2.6.0-test6 16 19.46 10.99% 9.435 2515.46 0.00000
2.6.0-test6-mm4 16 18.92 10.48% 9.650 2572.79 0.00000
2.6.0-test7 16 18.34 10.11% 10.014 2648.17 0.00000
2.6.0-test5 32 28.86 16.73% 12.003 1408.20 0.00000
2.6.0-test5-mm2 32 28.21 16.52% 12.099 1385.36 0.00000
2.6.0-test5-mm4 32 27.64 16.72% 12.706 1491.53 0.00000
2.6.0-test6 32 17.33 9.87% 20.872 5771.49 0.00005
2.6.0-test6-mm4 32 18.21 10.25% 19.659 4621.02 0.00000
2.6.0-test7 32 17.89 9.88% 20.121 4681.65 0.00000
2.6.0-test5 64 24.08 15.49% 27.936 1610.23 0.00000
2.6.0-test5-mm2 64 23.23 15.01% 27.942 1478.59 0.00000
2.6.0-test5-mm4 64 22.97 15.39% 29.020 1577.31 0.00000
2.6.0-test6 64 7.62 5.28% 95.479 16071.22 3.52058
2.6.0-test6-mm4 64 7.89 5.41% 91.677 12720.86 3.22466
2.6.0-test7 64 7.51 5.16% 96.373 14738.26 3.39680
2.6.0-test5 128 9.77 11.49% 141.347 1501.12 0.00000
2.6.0-test5-mm2 128 10.81 12.13% 121.146 1501.77 0.00000
2.6.0-test5-mm4 128 10.08 12.01% 134.006 1616.24 0.00000
2.6.0-test6 128 2.71 2.80% 534.963 33603.12 10.19945
2.6.0-test6-mm4 128 2.76 2.91% 524.334 34758.05 9.84798
2.6.0-test7 128 2.66 2.72% 543.918 33661.62 10.33516
2.6.0-test5 256 8.88 15.47% 327.283 2378.44 0.00000
2.6.0-test5-mm2 256 8.59 15.37% 339.514 1905.29 0.00000
2.6.0-test5-mm4 256 8.60 16.11% 338.642 1890.15 0.00000
2.6.0-test6 256 2.29 2.42% 1177.401 291951.14 11.55243
2.6.0-test6-mm4 256 2.32 2.50% 1158.512 343146.47 11.20739
2.6.0-test7 256 2.28 2.39% 1172.902 277837.90 11.42778
Random Reads reiserfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 0.88 0.95% 13.375 121.48 0.00000
2.6.0-test5-mm2 1 1.01 1.08% 11.543 382.20 0.00000
2.6.0-test5-mm4 1 0.86 1.11% 13.564 123.23 0.00000
2.6.0-test6 1 0.96 1.10% 12.246 127.82 0.00000
2.6.0-test6-mm4 1 1.02 1.16% 11.496 391.26 0.00000
2.6.0-test7 1 0.99 1.04% 11.863 111.96 0.00000
2.6.0-test5 8 4.59 5.28% 17.392 126.60 0.00000
2.6.0-test5-mm2 8 4.88 5.41% 15.978 105.27 0.00000
2.6.0-test5-mm4 8 4.64 5.92% 17.234 119.30 0.00000
2.6.0-test6 8 3.12 3.28% 27.584 373.26 0.00000
2.6.0-test6-mm4 8 3.80 4.12% 23.064 280.06 0.00000
2.6.0-test7 8 3.35 3.69% 24.895 337.23 0.00000
2.6.0-test5 16 4.80 4.33% 34.457 189.80 0.00000
2.6.0-test5-mm2 16 5.27 5.28% 30.882 160.61 0.00000
2.6.0-test5-mm4 16 5.22 5.44% 32.357 494.99 0.00000
2.6.0-test6 16 3.93 4.12% 43.768 728.55 0.00000
2.6.0-test6-mm4 16 3.96 3.79% 41.680 775.69 0.00000
2.6.0-test7 16 3.96 3.85% 42.797 705.06 0.00000
2.6.0-test5 32 5.45 5.07% 59.014 323.30 0.00000
2.6.0-test5-mm2 32 5.65 5.45% 57.673 313.96 0.00000
2.6.0-test5-mm4 32 4.47 4.84% 75.004 455.38 0.00000
2.6.0-test6 32 4.39 4.58% 77.879 1232.67 0.00000
2.6.0-test6-mm4 32 4.15 4.34% 80.355 1229.75 0.00000
2.6.0-test7 32 4.88 4.59% 67.103 1033.43 0.00000
2.6.0-test5 64 5.49 5.42% 108.692 991.66 0.00000
2.6.0-test5-mm2 64 5.10 5.60% 113.940 914.35 0.00000
2.6.0-test5-mm4 64 4.72 5.35% 126.505 1182.30 0.00000
2.6.0-test6 64 3.90 4.03% 165.855 2554.36 0.00000
2.6.0-test6-mm4 64 4.54 4.72% 140.927 2060.27 0.00000
2.6.0-test7 64 4.95 4.73% 129.947 2011.81 0.00000
2.6.0-test5 128 5.48 5.99% 221.089 2074.91 0.00000
2.6.0-test5-mm2 128 5.71 6.49% 208.548 1781.53 0.00000
2.6.0-test5-mm4 128 4.94 5.94% 240.781 2090.92 0.00000
2.6.0-test6 128 3.55 4.55% 367.628 5086.89 0.00000
2.6.0-test6-mm4 128 4.15 4.50% 314.516 3914.52 0.00000
2.6.0-test7 128 3.98 4.56% 328.297 4300.67 0.00000
2.6.0-test5 256 5.39 7.63% 455.466 2448.36 0.00000
2.6.0-test5-mm2 256 5.85 8.92% 415.957 2773.47 0.00000
2.6.0-test5-mm4 256 4.88 8.30% 498.235 2859.83 0.00000
2.6.0-test6 256 2.88 4.24% 861.877 8947.59 7.13542
2.6.0-test6-mm4 256 3.28 4.52% 752.088 7585.14 3.09895
2.6.0-test7 256 3.11 4.19% 794.027 8045.64 3.82812
Sequential Writes reiserf
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 55.30 90.83% 0.200 1317.87 0.00000
2.6.0-test5-mm2 1 55.08 92.70% 0.201 1512.52 0.00000
2.6.0-test5-mm4 1 54.62 92.30% 0.202 2146.88 0.00000
2.6.0-test6 1 55.09 91.14% 0.201 1243.21 0.00000
2.6.0-test6-mm4 1 54.46 91.70% 0.202 1725.38 0.00000
2.6.0-test7 1 55.08 91.27% 0.200 2381.55 0.00000
2.6.0-test5 8 27.22 99.99% 2.834 62892.07 0.02932
2.6.0-test5-mm2 8 27.84 100.4% 2.721 20683.58 0.03290
2.6.0-test5-mm4 8 28.18 96.26% 2.667 24253.31 0.03237
2.6.0-test6 8 28.88 102.9% 2.572 28097.64 0.03533
2.6.0-test6-mm4 8 26.57 92.64% 2.855 24274.27 0.05517
2.6.0-test7 8 28.60 104.5% 2.571 19829.03 0.02890
2.6.0-test5 16 30.58 102.6% 4.745 29788.03 0.07806
2.6.0-test5-mm2 16 28.94 101.7% 5.051 37491.58 0.07902
2.6.0-test5-mm4 16 28.45 97.86% 4.971 28952.01 0.08851
2.6.0-test6 16 27.13 98.05% 5.414 58605.69 0.08331
2.6.0-test6-mm4 16 27.04 89.13% 5.347 24565.13 0.08516
2.6.0-test7 16 29.19 98.45% 5.031 29653.75 0.08398
2.6.0-test5 32 27.88 104.6% 10.012 44186.44 0.11630
2.6.0-test5-mm2 32 27.15 104.7% 9.911 47326.17 0.11076
2.6.0-test5-mm4 32 27.57 102.6% 9.820 88173.93 0.10858
2.6.0-test6 32 27.80 100.8% 9.951 72577.05 0.11445
2.6.0-test6-mm4 32 25.80 97.74% 10.563 65400.15 0.11706
2.6.0-test7 32 28.65 105.3% 9.613 44890.58 0.11087
2.6.0-test5 64 27.12 109.5% 19.091 93664.16 0.15073
2.6.0-test5-mm2 64 25.80 117.2% 20.100 73309.25 0.18029
2.6.0-test5-mm4 64 25.64 112.9% 19.408 76139.55 0.16613
2.6.0-test6 64 26.30 112.0% 19.162 93011.66 0.16308
2.6.0-test6-mm4 64 24.15 104.3% 21.613 86562.64 0.18483
2.6.0-test7 64 25.10 107.7% 20.311 80093.32 0.16470
2.6.0-test5 128 23.52 119.4% 42.286 116297.76 0.39005
2.6.0-test5-mm2 128 24.88 127.8% 39.106 139584.40 0.35658
2.6.0-test5-mm4 128 24.28 127.9% 39.111 159921.42 0.37918
2.6.0-test6 128 24.79 122.4% 41.723 127648.00 0.39692
2.6.0-test6-mm4 128 24.48 125.2% 39.716 128936.67 0.38275
2.6.0-test7 128 24.29 122.0% 40.692 152191.29 0.35692
2.6.0-test5 256 19.14 163.4% 112.557 239944.32 1.00599
2.6.0-test5-mm2 256 23.21 153.9% 84.422 200276.08 0.75154
2.6.0-test5-mm4 256 23.45 161.4% 86.423 216520.39 0.71788
2.6.0-test6 256 23.14 168.2% 83.087 247571.24 0.75640
2.6.0-test6-mm4 256 20.87 136.1% 74.033 177291.47 0.66229
2.6.0-test7 256 23.11 142.0% 79.651 233921.91 0.66109
Random Writes reiserfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 2.41 2.77% 1.925 282.65 0.00000
2.6.0-test5-mm2 1 3.71 4.05% 0.222 18.72 0.00000
2.6.0-test5-mm4 1 2.29 2.87% 2.191 299.62 0.00000
2.6.0-test6 1 2.89 3.32% 1.040 91.48 0.00000
2.6.0-test6-mm4 1 3.60 3.66% 0.228 24.13 0.00000
2.6.0-test7 1 3.58 3.58% 0.347 30.23 0.00000
2.6.0-test5 8 3.50 4.12% 11.316 1348.95 0.00000
2.6.0-test5-mm2 8 3.74 5.07% 10.135 1133.62 0.00000
2.6.0-test5-mm4 8 3.54 4.73% 10.168 1285.63 0.00000
2.6.0-test6 8 3.26 4.02% 9.287 1182.99 0.00000
2.6.0-test6-mm4 8 3.39 4.53% 5.611 1106.14 0.00000
2.6.0-test7 8 3.29 3.77% 10.518 1191.14 0.00000
2.6.0-test5 16 3.44 4.05% 22.370 4353.76 0.10000
2.6.0-test5-mm2 16 3.54 3.72% 22.545 1478.64 0.00000
2.6.0-test5-mm4 16 3.61 4.71% 20.853 1777.27 0.00000
2.6.0-test6 16 3.60 4.11% 17.724 1421.63 0.00000
2.6.0-test6-mm4 16 3.37 3.72% 23.097 1422.10 0.00000
2.6.0-test7 16 3.46 4.40% 16.822 4822.00 0.02500
2.6.0-test5 32 3.68 4.57% 22.011 1260.71 0.00000
2.6.0-test5-mm2 32 3.60 4.50% 26.561 1836.41 0.00000
2.6.0-test5-mm4 32 3.55 4.32% 33.055 1460.77 0.00000
2.6.0-test6 32 3.41 4.07% 23.498 2013.53 0.00000
2.6.0-test6-mm4 32 3.42 4.36% 13.875 1214.49 0.00000
2.6.0-test7 32 3.42 4.04% 26.600 3437.82 0.02500
2.6.0-test5 64 3.62 5.09% 37.320 8700.40 0.90725
2.6.0-test5-mm2 64 3.81 6.11% 18.225 1814.21 0.00000
2.6.0-test5-mm4 64 3.37 5.09% 43.372 3280.58 0.00000
2.6.0-test6 64 2.96 3.93% 73.537 4484.06 0.00000
2.6.0-test6-mm4 64 3.29 4.98% 22.802 2206.78 0.00000
2.6.0-test7 64 3.42 5.00% 27.782 2453.44 0.00000
2.6.0-test5 128 3.73 5.92% 48.524 7848.01 1.10887
2.6.0-test5-mm2 128 3.39 5.91% 36.168 5651.07 0.25202
2.6.0-test5-mm4 128 3.55 6.74% 70.991 6856.35 1.94052
2.6.0-test6 128 3.16 5.58% 43.277 5286.77 0.05040
2.6.0-test6-mm4 128 3.17 5.25% 43.818 5534.13 0.30242
2.6.0-test7 128 3.26 5.26% 39.051 3037.10 0.00000
2.6.0-test5 256 3.96 8.27% 8.806 963.34 0.00000
2.6.0-test5-mm2 256 3.67 7.17% 20.000 1575.35 0.00000
2.6.0-test5-mm4 256 3.36 6.99% 114.310 4266.84 1.61458
2.6.0-test6 256 3.10 6.69% 132.312 8228.26 3.30729
2.6.0-test6-mm4 256 3.21 6.79% 88.532 8631.04 1.79687
2.6.0-test7 256 3.23 6.61% 48.643 6005.85 0.75521
Sequential Reads xfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 30.46 14.80% 0.382 718.97 0.00000
2.6.0-test5-mm2 1 28.43 14.31% 0.409 729.27 0.00000
2.6.0-test5-mm4 1 29.25 14.50% 0.398 751.10 0.00000
2.6.0-test6 1 29.25 14.52% 0.398 735.45 0.00000
2.6.0-test6-mm4 1 28.68 14.60% 0.406 924.52 0.00000
2.6.0-test7 1 29.20 14.46% 0.398 727.21 0.00000
2.6.0-test5 8 55.34 28.06% 1.652 831.66 0.00000
2.6.0-test5-mm2 8 54.19 28.05% 1.707 852.34 0.00000
2.6.0-test5-mm4 8 55.49 28.38% 1.591 1165.42 0.00000
2.6.0-test6 8 25.17 11.78% 3.562 1987.22 0.00000
2.6.0-test6-mm4 8 23.25 11.32% 3.989 1872.56 0.00000
2.6.0-test7 8 25.91 12.19% 3.443 1894.36 0.00000
2.6.0-test5 16 47.04 23.31% 3.903 846.86 0.00000
2.6.0-test5-mm2 16 45.50 23.01% 4.086 1116.23 0.00000
2.6.0-test5-mm4 16 47.00 23.78% 3.827 822.79 0.00000
2.6.0-test6 16 15.07 7.14% 12.090 3088.51 0.00000
2.6.0-test6-mm4 16 12.58 6.23% 14.655 3469.24 0.00000
2.6.0-test7 16 14.79 7.03% 11.808 3178.50 0.00000
2.6.0-test5 32 44.85 22.40% 7.699 1225.63 0.00000
2.6.0-test5-mm2 32 45.89 23.39% 7.624 1173.54 0.00000
2.6.0-test5-mm4 32 44.18 22.68% 7.940 1051.95 0.00000
2.6.0-test6 32 14.80 7.02% 24.021 5758.07 0.00005
2.6.0-test6-mm4 32 14.18 6.98% 25.241 5709.74 0.00000
2.6.0-test7 32 12.46 6.03% 29.086 5836.49 0.00029
2.6.0-test5 64 42.08 21.45% 16.269 1897.20 0.00000
2.6.0-test5-mm2 64 42.36 21.75% 15.948 1925.10 0.00000
2.6.0-test5-mm4 64 43.77 22.73% 15.396 1963.29 0.00000
2.6.0-test6 64 12.50 5.98% 53.864 10887.40 1.74727
2.6.0-test6-mm4 64 12.37 6.19% 55.962 10690.33 1.82176
2.6.0-test7 64 12.25 5.95% 56.211 10899.32 1.78017
2.6.0-test5 128 15.41 12.23% 87.001 2911.16 0.00000
2.6.0-test5-mm2 128 20.45 15.21% 61.549 2516.86 0.00000
2.6.0-test5-mm4 128 18.55 14.38% 68.572 2527.97 0.00000
2.6.0-test6 128 5.20 4.07% 251.782 18918.92 4.34985
2.6.0-test6-mm4 128 4.47 3.56% 290.774 19729.60 5.13649
2.6.0-test7 128 3.50 3.19% 385.879 20057.71 6.42790
2.6.0-test5 256 10.33 13.76% 225.449 6314.51 0.00167
2.6.0-test5-mm2 256 10.58 14.19% 206.025 9145.94 0.00219
2.6.0-test5-mm4 256 11.64 15.40% 175.294 11398.84 0.00677
2.6.0-test6 256 3.84 3.42% 539.606 972621.75 5.75909
2.6.0-test6-mm4 256 3.92 3.35% 527.548 867696.31 5.66021
2.6.0-test7 256 3.99 3.52% 511.822 865743.62 5.56827
Random Reads xfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 1.03 0.94% 11.357 107.82 0.00000
2.6.0-test5-mm2 1 1.03 1.02% 11.325 117.87 0.00000
2.6.0-test5-mm4 1 1.04 1.03% 11.261 107.89 0.00000
2.6.0-test6 1 1.03 0.92% 11.386 109.39 0.00000
2.6.0-test6-mm4 1 1.00 1.04% 11.659 119.70 0.00000
2.6.0-test7 1 1.05 0.92% 11.161 96.11 0.00000
2.6.0-test5 8 5.42 5.19% 16.257 124.33 0.00000
2.6.0-test5-mm2 8 5.36 5.79% 16.591 120.16 0.00000
2.6.0-test5-mm4 8 4.67 4.80% 16.448 104.77 0.00000
2.6.0-test6 8 2.08 1.68% 38.061 775.61 0.00000
2.6.0-test6-mm4 8 2.16 1.95% 37.981 686.42 0.00000
2.6.0-test7 8 1.88 1.64% 42.173 683.55 0.00000
2.6.0-test5 16 5.20 4.62% 34.058 146.53 0.00000
2.6.0-test5-mm2 16 5.21 4.88% 34.306 175.88 0.00000
2.6.0-test5-mm4 16 5.18 4.84% 32.751 135.47 0.00000
2.6.0-test6 16 1.86 1.44% 85.898 1244.64 0.00000
2.6.0-test6-mm4 16 1.95 1.64% 80.971 1175.91 0.00000
2.6.0-test7 16 2.35 1.98% 67.271 1001.72 0.00000
2.6.0-test5 32 4.91 3.98% 68.586 265.91 0.00000
2.6.0-test5-mm2 32 5.43 4.74% 63.283 279.95 0.00000
2.6.0-test5-mm4 32 4.69 4.28% 68.452 638.24 0.00000
2.6.0-test6 32 2.92 2.41% 113.627 1626.83 0.00000
2.6.0-test6-mm4 32 3.14 2.92% 105.714 1451.06 0.00000
2.6.0-test7 32 2.43 1.91% 134.936 1864.81 0.00000
2.6.0-test5 64 4.76 4.02% 136.805 879.54 0.00000
2.6.0-test5-mm2 64 4.66 4.68% 136.498 871.90 0.00000
2.6.0-test5-mm4 64 4.81 4.75% 131.041 994.54 0.00000
2.6.0-test6 64 1.45 1.25% 425.661 5262.95 0.02520
2.6.0-test6-mm4 64 1.66 1.40% 382.569 4759.16 0.00000
2.6.0-test7 64 1.41 1.21% 444.027 5820.44 0.30242
2.6.0-test5 128 4.70 4.50% 248.562 1396.77 0.00000
2.6.0-test5-mm2 128 5.29 5.72% 206.313 1395.33 0.00000
2.6.0-test5-mm4 128 5.16 4.95% 209.001 1485.37 0.00000
2.6.0-test6 128 1.65 1.57% 753.959 9498.63 9.02218
2.6.0-test6-mm4 128 1.47 1.43% 833.261 9951.50 11.54233
2.6.0-test7 128 1.23 1.06% 969.453 10793.66 18.04436
2.6.0-test5 256 4.82 6.71% 469.459 2140.97 0.00000
2.6.0-test5-mm2 256 4.60 6.06% 477.682 1934.27 0.00000
2.6.0-test5-mm4 256 5.01 6.60% 402.622 1881.24 0.00000
2.6.0-test6 256 0.90 0.99% 2204.619 32794.62 42.68229
2.6.0-test6-mm4 256 0.88 0.98% 2165.705 33928.64 41.09376
2.6.0-test7 256 0.92 1.05% 2113.981 33637.27 39.11458
Sequential Writes xfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 56.44 46.30% 0.182 68379.68 0.00030
2.6.0-test5-mm2 1 56.33 47.39% 0.186 68474.74 0.00030
2.6.0-test5-mm4 1 56.30 46.74% 0.182 68268.99 0.00030
2.6.0-test6 1 56.24 46.08% 0.177 68501.77 0.00030
2.6.0-test6-mm4 1 56.17 47.57% 0.175 68404.17 0.00030
2.6.0-test7 1 56.26 46.03% 0.171 68451.18 0.00030
2.6.0-test5 8 57.54 65.47% 0.991 166706.97 0.00424
2.6.0-test5-mm2 8 57.70 70.44% 1.099 211950.97 0.00420
2.6.0-test5-mm4 8 57.23 66.09% 1.042 205671.61 0.00415
2.6.0-test6 8 57.19 69.31% 1.058 185132.73 0.00410
2.6.0-test6-mm4 8 57.46 66.82% 1.100 178544.18 0.00486
2.6.0-test7 8 57.24 64.93% 1.096 207662.87 0.00481
2.6.0-test5 16 57.26 71.20% 2.069 186551.11 0.01030
2.6.0-test5-mm2 16 57.26 72.02% 2.079 187786.90 0.01059
2.6.0-test5-mm4 16 57.38 68.91% 1.999 193908.75 0.00958
2.6.0-test6 16 57.24 70.72% 2.072 185124.29 0.01092
2.6.0-test6-mm4 16 57.37 68.99% 2.080 208981.25 0.01045
2.6.0-test7 16 57.17 68.79% 2.051 181396.29 0.01135
2.6.0-test5 32 57.29 75.00% 4.076 211670.36 0.01840
2.6.0-test5-mm2 32 57.34 73.21% 3.768 213556.33 0.01560
2.6.0-test5-mm4 32 57.27 75.59% 4.007 199887.88 0.01708
2.6.0-test6 32 57.12 73.89% 4.007 198273.60 0.01750
2.6.0-test6-mm4 32 57.04 73.27% 4.034 181071.17 0.01688
2.6.0-test7 32 57.04 72.85% 4.066 193571.15 0.02007
2.6.0-test5 64 57.13 83.01% 8.186 336901.61 0.03567
2.6.0-test5-mm2 64 56.93 83.63% 8.062 340355.98 0.03066
2.6.0-test5-mm4 64 56.33 82.10% 7.834 325305.64 0.02895
2.6.0-test6 64 56.48 80.10% 7.829 306781.15 0.03157
2.6.0-test6-mm4 64 56.87 81.05% 7.808 257430.27 0.03662
2.6.0-test7 64 56.54 77.01% 7.746 347269.10 0.02880
2.6.0-test5 128 56.58 86.45% 15.451 330514.59 0.06585
2.6.0-test5-mm2 128 56.78 82.31% 14.164 364774.73 0.05298
2.6.0-test5-mm4 128 55.57 82.39% 14.088 350137.61 0.05364
2.6.0-test6 128 56.08 81.22% 13.917 377412.59 0.04669
2.6.0-test6-mm4 128 56.43 79.05% 14.332 357252.17 0.05203
2.6.0-test7 128 56.28 77.14% 14.465 351042.94 0.05546
2.6.0-test5 256 56.13 83.69% 19.836 383235.17 0.07386
2.6.0-test5-mm2 256 56.19 80.90% 16.880 294598.89 0.07434
2.6.0-test5-mm4 256 56.09 87.58% 13.343 308894.90 0.05746
2.6.0-test6 256 55.39 80.57% 18.269 350145.38 0.07200
2.6.0-test6-mm4 256 55.67 81.98% 17.338 337290.04 0.06890
2.6.0-test7 256 55.49 82.29% 15.776 351541.33 0.06071
Random Writes xfs
Num Avg Maximum Lat%
Kernel Thr Rate (CPU%) Latency Latency >2s
------------------------- --- --------------------------------------------
2.6.0-test5 1 3.80 3.61% 0.074 0.46 0.00000
2.6.0-test5-mm2 1 3.86 3.77% 0.077 0.42 0.00000
2.6.0-test5-mm4 1 3.77 3.65% 0.078 1.42 0.00000
2.6.0-test6 1 3.76 3.53% 0.075 1.43 0.00000
2.6.0-test6-mm4 1 3.71 3.64% 0.078 0.68 0.00000
2.6.0-test7 1 3.76 3.53% 0.075 0.87 0.00000
2.6.0-test5 8 3.97 6.78% 0.202 95.59 0.00000
2.6.0-test5-mm2 8 3.91 6.75% 0.181 75.97 0.00000
2.6.0-test5-mm4 8 3.87 6.57% 0.191 81.43 0.00000
2.6.0-test6 8 3.86 6.63% 0.165 25.62 0.00000
2.6.0-test6-mm4 8 3.82 6.33% 0.166 39.50 0.00000
2.6.0-test7 8 3.92 6.44% 0.181 74.95 0.00000
2.6.0-test5 16 3.95 6.81% 0.168 32.20 0.00000
2.6.0-test5-mm2 16 3.94 6.99% 0.175 48.13 0.00000
2.6.0-test5-mm4 16 3.82 6.53% 0.186 87.83 0.00000
2.6.0-test6 16 3.76 6.58% 0.168 25.32 0.00000
2.6.0-test6-mm4 16 3.81 6.62% 0.193 83.97 0.00000
2.6.0-test7 16 3.89 6.57% 0.181 92.60 0.00000
2.6.0-test5 32 3.85 6.64% 0.159 3.38 0.00000
2.6.0-test5-mm2 32 3.80 6.75% 0.175 59.86 0.00000
2.6.0-test5-mm4 32 3.72 6.60% 0.180 55.04 0.00000
2.6.0-test6 32 3.80 6.73% 0.175 56.73 0.00000
2.6.0-test6-mm4 32 3.81 6.61% 0.171 55.65 0.00000
2.6.0-test7 32 3.60 5.97% 0.155 3.09 0.00000
2.6.0-test5 64 3.68 6.56% 0.161 3.90 0.00000
2.6.0-test5-mm2 64 3.60 6.51% 0.163 6.67 0.00000
2.6.0-test5-mm4 64 3.71 6.58% 0.166 5.44 0.00000
2.6.0-test6 64 3.53 6.32% 0.160 4.14 0.00000
2.6.0-test6-mm4 64 3.54 6.21% 0.185 106.22 0.00000
2.6.0-test7 64 3.61 6.15% 0.156 4.34 0.00000
2.6.0-test5 128 3.32 6.23% 0.166 4.42 0.00000
2.6.0-test5-mm2 128 3.64 6.70% 0.163 2.98 0.00000
2.6.0-test5-mm4 128 3.71 6.85% 0.169 5.37 0.00000
2.6.0-test6 128 3.44 6.25% 0.174 63.35 0.00000
2.6.0-test6-mm4 128 3.49 6.36% 0.174 64.77 0.00000
2.6.0-test7 128 3.43 6.15% 0.159 4.05 0.00000
2.6.0-test5 256 3.35 6.56% 0.161 3.65 0.00000
2.6.0-test5-mm2 256 3.29 6.41% 0.164 3.68 0.00000
2.6.0-test5-mm4 256 3.27 6.35% 0.166 4.84 0.00000
2.6.0-test6 256 2.97 5.80% 0.171 47.92 0.00000
2.6.0-test6-mm4 256 3.06 5.93% 0.159 3.02 0.00000
2.6.0-test7 256 3.11 5.88% 0.159 3.91 0.00000
dbench did not regress.
dbench-1.3 reiserfs 192 processes Average +/- MB/sec
2.6.0-test7 77.3 7.1
2.6.0-test6-mm4 76.6 7.2
2.6.0-test5-mm2 76.1 6.8
2.6.0-test6 73.0 5.1
2.6.0-test5-mm4 72.5 6.6
2.6.0-test5 65.0 1.2
dbench-1.3 reiserfs 64 processes Average +/- MB/sec
2.6.0-test5-mm4 111.4 9.1
2.6.0-test6-mm4 109.6 7.5
2.6.0-test6 108.8 4.2
2.6.0-test5-mm2 104.4 8.1
2.6.0-test7 102.3 8.1
2.6.0-test5 100.6 4.7
dbench-1.3 xfs 192 processes Average +/- MB/sec
2.6.0-test5 112.5 4.0
2.6.0-test5-mm2 110.8 1.6
2.6.0-test6-mm4 110.7 3.4
2.6.0-test5-mm4 109.7 2.0
2.6.0-test7 109.0 5.1
2.6.0-test6 108.5 6.1
dbench-1.3 xfs 64 processes Average +/- MB/sec
2.6.0-test5 143.7 0.9
2.6.0-test7 143.6 2.1
2.6.0-test5-mm2 141.5 1.6
2.6.0-test6 141.4 0.5
2.6.0-test6-mm4 141.2 1.5
2.6.0-test5-mm4 140.0 1.4
dbench-1.3 ext2 192 processes Average +/- MB/sec
2.6.0-test6 209.5 11.8
2.6.0-test7 205.2 20.8
2.6.0-test5-mm4 200.8 6.2
2.6.0-test5-mm2 200.0 10.3
2.6.0-test5 198.1 19.5
2.6.0-test6-mm4 193.4 12.0
dbench-1.3 ext2 64 processes Average +/- MB/sec
2.6.0-test7 256.8 0.6
2.6.0-test6 256.7 1.6
2.6.0-test5 256.2 1.6
2.6.0-test5-mm4 255.5 1.4
2.6.0-test5-mm2 255.3 0.9
2.6.0-test6-mm4 255.3 1.7
dbench-1.3 jfs 192 processes Average +/- MB/sec
2.6.0-test5-mm4 11.3 0.5
2.6.0-test7 11.1 0.2
2.6.0-test6-mm4 11.0 0.3
dbench-1.3 jfs 64 processes Average +/- MB/sec
2.6.0-test5-mm2 20.0 2.9
2.6.0-test5 19.5 2.4
2.6.0-test5-mm4 18.6 2.4
2.6.0-test7 18.5 1.1
2.6.0-test6-mm4 18.1 1.6
dbench-1.3 ext3 192 processes Average +/- MB/sec
2.6.0-test7 110.8 23.1
2.6.0-test5-mm2 107.5 39.7
2.6.0-test6 103.9 39.0
2.6.0-test6-mm4 99.7 38.7
2.6.0-test5-mm4 82.6 35.5
2.6.0-test5 73.6 15.1
dbench-1.3 ext3 64 processes Average +/- MB/sec
2.6.0-test7 177.9 3.4
2.6.0-test6 175.1 3.6
2.6.0-test5 174.5 4.2
2.6.0-test6-mm4 174.2 2.5
2.6.0-test5-mm2 172.8 4.0
2.6.0-test5-mm4 172.6 2.3
--
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html
^ permalink raw reply [flat|nested] 25+ messages in thread
end of thread, other threads:[~2003-10-29 0:35 UTC | newest]
Thread overview: 25+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-10-29 0:38 [BENCHMARK] I/O regression after 2.6.0-test5 rwhron
-- strict thread matches above, loose matches on Subject: below --
2003-10-27 23:33 John Hawkes
2003-10-26 10:38 rwhron
2003-10-27 23:49 ` Dave Olien
2003-10-24 23:43 rwhron
2003-10-24 12:46 rwhron
2003-10-24 0:10 rwhron
2003-10-21 13:05 rwhron
2003-10-21 15:57 ` Nick Piggin
2003-10-21 21:07 ` venom
2003-10-22 9:54 ` Nick Piggin
2003-10-22 18:30 ` Dave Olien
2003-10-22 19:08 ` Dave Olien
2003-10-23 2:24 ` Nick Piggin
2003-10-23 3:48 ` Nick Piggin
2003-10-23 20:35 ` Dave Olien
2003-10-23 23:07 ` Nick Piggin
2003-10-20 0:37 rwhron
2003-10-20 1:35 ` Nick Piggin
2003-10-20 4:52 ` Andrew Morton
2003-10-20 8:12 ` Nick Piggin
2003-10-20 4:51 ` Dave Olien
2003-10-20 7:55 ` venom
2003-10-21 4:49 ` Nick Piggin
2003-10-24 13:52 ` Magnus Andersson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).