* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-13 16:47 Paolo Ciarrocchi
2002-10-13 21:02 ` Bill Davidsen
0 siblings, 1 reply; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-13 16:47 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
From: Bill Davidsen <davidsen@tmr.com>
[...]
> run this version I'd like to see the result. I believe I had to use the
> "-l" patch option to ignore blank mismatches to get this to work, and I've
> cleaned up another mailing funny as well.
Hi Bill,
here the results agains 2.5.41-mm2C (2.5.41-mm2 + Con patch "vmscan.c")
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.648 114.508 113.707 113.861 0.000 1.000
smallwrite 116.054 180.420 117.924 130.525 0.028 1.146
largewrite 114.019 179.770 120.451 134.021 0.028 1.177
cpuload 106.590 162.893 107.075 118.080 0.025 1.037
spawnload 106.574 164.898 107.490 118.671 0.026 1.042
8ctx-mem 7767.843 16917.625 8994.265 10906.788 3.844 95.790
2ctx-mem 6515.450 18273.101 10344.575 11217.755 4.822 98.521
8ctx-mem and 2ctx-mem show "bad" performance.
Do you think is it possible to apply the patch on the top of 2.5.42-mm2 ?
Ciao,
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
2002-10-13 16:47 Re:Benchmark results from resp1 trivial response time test Paolo Ciarrocchi
@ 2002-10-13 21:02 ` Bill Davidsen
0 siblings, 0 replies; 11+ messages in thread
From: Bill Davidsen @ 2002-10-13 21:02 UTC (permalink / raw)
To: Paolo Ciarrocchi; +Cc: linux-kernel
On Mon, 14 Oct 2002, Paolo Ciarrocchi wrote:
> From: Bill Davidsen <davidsen@tmr.com>
> [...]
> > run this version I'd like to see the result. I believe I had to use the
> > "-l" patch option to ignore blank mismatches to get this to work, and I've
> > cleaned up another mailing funny as well.
>
> Hi Bill,
> here the results agains 2.5.41-mm2C (2.5.41-mm2 + Con patch "vmscan.c")
>
> Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
>
> _____________ delay ms. ____________
> Test low high median average S.D. ratio
> noload 113.648 114.508 113.707 113.861 0.000 1.000
> smallwrite 116.054 180.420 117.924 130.525 0.028 1.146
> largewrite 114.019 179.770 120.451 134.021 0.028 1.177
> cpuload 106.590 162.893 107.075 118.080 0.025 1.037
> spawnload 106.574 164.898 107.490 118.671 0.026 1.042
> 8ctx-mem 7767.843 16917.625 8994.265 10906.788 3.844 95.790
> 2ctx-mem 6515.450 18273.101 10344.575 11217.755 4.822 98.521
>
> 8ctx-mem and 2ctx-mem show "bad" performance.
> Do you think is it possible to apply the patch on the top of 2.5.42-mm2 ?
I haven't tried it yet, but I'm interested in your result, since my
2.5.41-mm2v result was actually better then plain -mm2. I am just building
some new test stuff on an SMP machine so I can compare uni and SMP
performance under load, and I'll look at 2.5.42 tonight or tomorrow.
My reference machine was a 96MB machine, if you're really curious about
this you could boot with mem=96m (or 128m) and rerun the test. In any case
I have the kids tonight, if I get some time I'll try it, otherwise
tomorrow.
--
bill davidsen <davidsen@tmr.com>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-17 20:16 Paolo Ciarrocchi
0 siblings, 0 replies; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-17 20:16 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel, akpm
Updated results.
Kernel version: 2.4.19
Starting 1 CPU run with 251 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 119.653 124.207 119.786 120.644 0.002 1.000
smallwrite 121.051 190.696 144.469 152.062 0.031 1.260
largewrite 117.412 11871.027 1113.563 3362.302 4.870 27.870
cpuload 119.691 313.702 178.513 190.324 0.081 1.578
spawnload 114.624 243.985 119.747 154.604 0.057 1.281
8ctx-mem 652.737 7564.493 1234.938 2527.273 2.904 20.948
2ctx-mem 126.914 8428.021 620.418 2941.043 3.890 24.378
Kernel version: 2.5.42
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 113.948 113.737 113.774 0.000 1.000
smallwrite 115.352 189.983 119.500 141.645 0.036 1.245
largewrite 19012.080 45528.170 30311.890 31472.632 9.646 276.625
cpuload 103.328 1267.116 103.971 336.462 0.520 2.957
spawnload 105.196 167.400 105.787 117.983 0.028 1.037
8ctx-mem 121.166 8616.232 126.489 1824.054 3.797 16.032
2ctx-mem 115.162 10291.470 119.560 2152.930 4.550 18.923
Kernel version: 2.5.42-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.658 114.500 113.737 113.866 0.000 1.000
smallwrite 117.220 245.110 119.017 144.541 0.056 1.269
largewrite 114.841 181.420 118.746 131.568 0.028 1.155
cpuload 104.583 159.080 104.890 115.694 0.024 1.016
spawnload 108.790 166.141 109.655 120.698 0.025 1.060
8ctx-mem 114.430 8484.873 784.141 2066.476 3.604 18.148
2ctx-mem 121.388 8453.018 1414.443 2383.773 3.459 20.935
Kernel version: 2.5.42-mm3
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.616 114.280 113.714 113.823 0.000 1.000
smallwrite 119.452 188.512 123.024 138.799 0.029 1.219
largewrite 117.059 177.488 119.743 131.807 0.026 1.158
cpuload 104.678 165.664 104.837 117.055 0.027 1.028
spawnload 108.795 162.441 108.809 119.582 0.024 1.051
8ctx-mem 1056.455 10130.557 2031.861 3478.685 3.810 30.562
2ctx-mem 113.234 10068.534 1518.290 2861.645 4.073 25.141
Kernel version: 2.5.43
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.596 114.094 113.697 113.779 0.000 1.000
smallwrite 114.992 181.439 121.280 139.729 0.032 1.228
largewrite 113.795 262.061 121.018 158.819 0.064 1.396
cpuload 105.404 161.821 105.760 117.027 0.025 1.029
spawnload 107.522 166.259 107.645 119.364 0.026 1.049
8ctx-mem 1771.175 9863.631 2593.279 4018.103 3.316 35.315
2ctx-mem 1421.036 10475.948 1996.562 3812.757 3.785 33.510
Kernel version: 2.5.43-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.627 114.172 113.748 113.794 0.000 1.000
smallwrite 116.479 184.859 144.951 147.506 0.032 1.296
largewrite 114.025 986.386 173.816 325.635 0.373 2.862
cpuload 104.916 159.896 105.425 116.283 0.024 1.022
spawnload 104.647 171.033 104.931 118.055 0.030 1.037
8ctx-mem 118.156 9500.830 301.199 2132.289 4.124 18.738
2ctx-mem 113.554 10007.751 636.093 2542.432 4.231 22.342
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-15 21:03 Paolo Ciarrocchi
0 siblings, 0 replies; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-15 21:03 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
Hi all,
updated results:
Kernel version: 2.4.19
Starting 1 CPU run with 251 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 119.653 124.207 119.786 120.644 0.002 1.000
smallwrite 121.051 190.696 144.469 152.062 0.031 1.260
largewrite 117.412 11871.027 1113.563 3362.302 4.870 27.870
cpuload 119.691 313.702 178.513 190.324 0.081 1.578
spawnload 114.624 243.985 119.747 154.604 0.057 1.281
8ctx-mem 652.737 7564.493 1234.938 2527.273 2.904 20.948
2ctx-mem 126.914 8428.021 620.418 2941.043 3.890 24.378
Kernel version: 2.5.41-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 114.413 113.713 113.856 0.000 1.000
smallwrite 117.001 228.536 119.025 141.028 0.049 1.239
largewrite 116.102 1681.086 216.776 499.422 0.667 4.386
cpuload 105.461 160.513 105.834 116.661 0.025 1.025
spawnload 108.506 166.367 108.863 120.305 0.026 1.057
8ctx-mem 1789.834 9155.800 3790.198 4547.118 2.831 39.938
2ctx-mem 2290.391 10661.156 4810.641 5040.725 3.358 44.273
Kernel version: 2.5.41-mm2C
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.648 114.508 113.707 113.861 0.000 1.000
smallwrite 116.054 180.420 117.924 130.525 0.028 1.146
largewrite 114.019 179.770 120.451 134.021 0.028 1.177
cpuload 106.590 162.893 107.075 118.080 0.025 1.037
spawnload 106.574 164.898 107.490 118.671 0.026 1.042
8ctx-mem 7767.843 16917.625 8994.265 10906.788 3.844 95.790
2ctx-mem 6515.450 18273.101 10344.575 11217.755 4.822 98.521
Kernel version: 2.5.41-mm2C@128MiB
Starting 1 CPU run with 123 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.550 114.451 113.763 113.857 0.000 1.000
smallwrite 118.926 182.074 123.044 139.160 0.028 1.222
largewrite 119.864 186.550 139.593 148.957 0.027 1.308
cpuload 105.468 165.026 105.922 117.648 0.026 1.033
spawnload 106.408 162.414 106.778 117.861 0.025 1.035
8ctx-mem 7264.508 12232.155 8106.800 8836.428 1.965 77.610
2ctx-mem 7447.326 23237.534 11601.912 13325.664 6.527 117.038
Kernel version: 2.5.41
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.607 114.030 113.774 113.783 0.000 1.000
smallwrite 117.882 526.275 121.038 208.868 0.178 1.836
largewrite 1674.216 88428.871 14794.752 27750.852 34.851 243.892
cpuload 104.225 158.400 104.478 115.269 0.024 1.013
spawnload 105.933 166.818 106.682 118.452 0.027 1.041
8ctx-mem 116.458 8893.645 120.670 1875.275 3.923 16.481
2ctx-mem 116.847 10174.152 121.309 2130.303 4.497 18.722
Kernel version: 2.5.42-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.658 114.500 113.737 113.866 0.000 1.000
smallwrite 117.220 245.110 119.017 144.541 0.056 1.269
largewrite 114.841 181.420 118.746 131.568 0.028 1.155
cpuload 104.583 159.080 104.890 115.694 0.024 1.016
spawnload 108.790 166.141 109.655 120.698 0.025 1.060
8ctx-mem 114.430 8484.873 784.141 2066.476 3.604 18.148
2ctx-mem 121.388 8453.018 1414.443 2383.773 3.459 20.935
Kernel version: 2.5.42-mm3
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.616 114.280 113.714 113.823 0.000 1.000
smallwrite 119.452 188.512 123.024 138.799 0.029 1.219
largewrite 117.059 177.488 119.743 131.807 0.026 1.158
cpuload 104.678 165.664 104.837 117.055 0.027 1.028
spawnload 108.795 162.441 108.809 119.582 0.024 1.051
8ctx-mem 1056.455 10130.557 2031.861 3478.685 3.810 30.562
2ctx-mem 113.234 10068.534 1518.290 2861.645 4.073 25.141
Kernel version: 2.5.42
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 113.948 113.737 113.774 0.000 1.000
smallwrite 115.352 189.983 119.500 141.645 0.036 1.245
largewrite 19012.080 45528.170 30311.890 31472.632 9.646 276.625
cpuload 103.328 1267.116 103.971 336.462 0.520 2.957
spawnload 105.196 167.400 105.787 117.983 0.028 1.037
8ctx-mem 121.166 8616.232 126.489 1824.054 3.797 16.032
2ctx-mem 115.162 10291.470 119.560 2152.930 4.550 18.923
Ciao,
Paolo
P.S Andrew, do you like test and results ?
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-15 19:15 Paolo Ciarrocchi
0 siblings, 0 replies; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-15 19:15 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
From: Bill Davidsen <davidsen@tmr.com>
> On Tue, 15 Oct 2002, Paolo Ciarrocchi wrote:
>
> > Hi Bill,
> > I'm back with the results of others tests, here all my results:
>
> Thanks, I'll cip them in the response, but this test sure does make some
> kernels unhappy, doesn't it? And the sad truth is that this isn't
> artifact, if you get a similar "real load" on the machine the response
> will be really unusable in real life.
Yeah, but it also means that 2.5.42-mm2 is the winner, and I have the same feeling just using it, so I think we'll see a 2.5.43 very responsive ;-)
Did you try 2.5.42-mm3 ?
Ciao,
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
2002-10-14 21:30 Paolo Ciarrocchi
@ 2002-10-15 1:31 ` Bill Davidsen
0 siblings, 0 replies; 11+ messages in thread
From: Bill Davidsen @ 2002-10-15 1:31 UTC (permalink / raw)
To: Paolo Ciarrocchi; +Cc: linux-kernel
On Tue, 15 Oct 2002, Paolo Ciarrocchi wrote:
> Hi Bill,
> I'm back with the results of others tests, here all my results:
Thanks, I'll cip them in the response, but this test sure does make some
kernels unhappy, doesn't it? And the sad truth is that this isn't
artifact, if you get a similar "real load" on the machine the response
will be really unusable in real life.
> I post the script I use to get this summary as well,
> do you think it is usefull ?
Sure, it's very like the one I used to generate the results on the web
site, other than I used "version=${filename%.out}" to avoid a cut or sed
process, and I put my redirect after the "done" as "done >>summary.txt" so
I didn't have to put it on every line. Matter of style, and I'm really
lazy about typing anything I don't need to ;-)
> #!/bin/bash
> out=(`ls *.out|sort`)
> total=`echo ${out[@]}|wc -w`
> # echo $total
> > summary.txt
> for i in `seq 0 1 $[total-1]`
> do
> kernel_version=`echo ${out[i]}|cut -d "." --fields=1-3`
> echo -e "\t\tKernel version: $kernel_version" >> summary.txt
> grep '^ ' ${out[i]} >> summary.txt
> echo >> summary.txt
> done;
I'm going to replace S.D. with the ratio of the median values in the near
future, I find that more useful, since one really bad response can skew
the ratio of average. Of course the other values are there to avoid
confusion, but t makes a nicer number for the "look at one value" folks.
--
bill davidsen <davidsen@tmr.com>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-14 21:30 Paolo Ciarrocchi
2002-10-15 1:31 ` Bill Davidsen
0 siblings, 1 reply; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-14 21:30 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
Hi Bill,
I'm back with the results of others tests, here all my results:
Kernel version: 2.4.19
Starting 1 CPU run with 251 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 119.653 124.207 119.786 120.644 0.002 1.000
smallwrite 121.051 190.696 144.469 152.062 0.031 1.260
largewrite 117.412 11871.027 1113.563 3362.302 4.870 27.870
cpuload 119.691 313.702 178.513 190.324 0.081 1.578
spawnload 114.624 243.985 119.747 154.604 0.057 1.281
8ctx-mem 652.737 7564.493 1234.938 2527.273 2.904 20.948
2ctx-mem 126.914 8428.021 620.418 2941.043 3.890 24.378
Kernel version: 2.5.41-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 114.413 113.713 113.856 0.000 1.000
smallwrite 117.001 228.536 119.025 141.028 0.049 1.239
largewrite 116.102 1681.086 216.776 499.422 0.667 4.386
cpuload 105.461 160.513 105.834 116.661 0.025 1.025
spawnload 108.506 166.367 108.863 120.305 0.026 1.057
8ctx-mem 1789.834 9155.800 3790.198 4547.118 2.831 39.938
2ctx-mem 2290.391 10661.156 4810.641 5040.725 3.358 44.273
Kernel version: 2.5.41-mm2C
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.648 114.508 113.707 113.861 0.000 1.000
smallwrite 116.054 180.420 117.924 130.525 0.028 1.146
largewrite 114.019 179.770 120.451 134.021 0.028 1.177
cpuload 106.590 162.893 107.075 118.080 0.025 1.037
spawnload 106.574 164.898 107.490 118.671 0.026 1.042
8ctx-mem 7767.843 16917.625 8994.265 10906.788 3.844 95.790
2ctx-mem 6515.450 18273.101 10344.575 11217.755 4.822 98.521
Kernel version: 2.5.41-mm2C@128MiB
Starting 1 CPU run with 123 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.550 114.451 113.763 113.857 0.000 1.000
smallwrite 118.926 182.074 123.044 139.160 0.028 1.222
largewrite 119.864 186.550 139.593 148.957 0.027 1.308
cpuload 105.468 165.026 105.922 117.648 0.026 1.033
spawnload 106.408 162.414 106.778 117.861 0.025 1.035
8ctx-mem 7264.508 12232.155 8106.800 8836.428 1.965 77.610
2ctx-mem 7447.326 23237.534 11601.912 13325.664 6.527 117.038
Kernel version: 2.5.41
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.607 114.030 113.774 113.783 0.000 1.000
smallwrite 117.882 526.275 121.038 208.868 0.178 1.836
largewrite 1674.216 88428.871 14794.752 27750.852 34.851 243.892
cpuload 104.225 158.400 104.478 115.269 0.024 1.013
spawnload 105.933 166.818 106.682 118.452 0.027 1.041
8ctx-mem 116.458 8893.645 120.670 1875.275 3.923 16.481
2ctx-mem 116.847 10174.152 121.309 2130.303 4.497 18.722
Kernel version: 2.5.42-mm2
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.658 114.500 113.737 113.866 0.000 1.000
smallwrite 117.220 245.110 119.017 144.541 0.056 1.269
largewrite 114.841 181.420 118.746 131.568 0.028 1.155
cpuload 104.583 159.080 104.890 115.694 0.024 1.016
spawnload 108.790 166.141 109.655 120.698 0.025 1.060
8ctx-mem 114.430 8484.873 784.141 2066.476 3.604 18.148
2ctx-mem 121.388 8453.018 1414.443 2383.773 3.459 20.935
Kernel version: 2.5.42
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 113.948 113.737 113.774 0.000 1.000
smallwrite 115.352 189.983 119.500 141.645 0.036 1.245
largewrite 19012.080 45528.170 30311.890 31472.632 9.646 276.625
cpuload 103.328 1267.116 103.971 336.462 0.520 2.957
spawnload 105.196 167.400 105.787 117.983 0.028 1.037
8ctx-mem 121.166 8616.232 126.489 1824.054 3.797 16.032
2ctx-mem 115.162 10291.470 119.560 2152.930 4.550 18.923
I post the script I use to get this summary as well,
do you think it is usefull ?
#!/bin/bash
out=(`ls *.out|sort`)
total=`echo ${out[@]}|wc -w`
# echo $total
> summary.txt
for i in `seq 0 1 $[total-1]`
do
kernel_version=`echo ${out[i]}|cut -d "." --fields=1-3`
echo -e "\t\tKernel version: $kernel_version" >> summary.txt
grep '^ ' ${out[i]} >> summary.txt
echo >> summary.txt
done;
Ciao,
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-14 7:38 Paolo Ciarrocchi
0 siblings, 0 replies; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-14 7:38 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
From: Bill Davidsen <davidsen@tmr.com>
[...]
> > Do you think is it possible to apply the patch on the top of 2.5.42-mm2 ?
>
> I haven't tried it yet, but I'm interested in your result, since my
> 2.5.41-mm2v result was actually better then plain -mm2. I am just building
> some new test stuff on an SMP machine so I can compare uni and SMP
> performance under load, and I'll look at 2.5.42 tonight or tomorrow.
Well, I did't run the test against 2.5.41-mm2, I'll do it just to compare the result with 2.5.42-mm2+vmscan patch.
> My reference machine was a 96MB machine, if you're really curious about
> this you could boot with mem=96m (or 128m) and rerun the test. In any case
> I have the kids tonight, if I get some time I'll try it, otherwise
> tomorrow.
Ok, I'll try to find the time to run the test against 2.5.42 and 96 or 128MiB of Ram.
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-13 15:25 Paolo Ciarrocchi
0 siblings, 0 replies; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-13 15:25 UTC (permalink / raw)
To: davidsen; +Cc: linux-kernel
From: Bill Davidsen <davidsen@tmr.com>
> On Sun, 13 Oct 2002, Paolo Ciarrocchi wrote:
>
> > Hi David,
> thanks for the quick results, but it's Bill...
Ops... sorry ;-)
> > I think your benchmark is very intersting.
> > Here goes my results:
>
> > It seems that 2.5.42-mm2 is the "winner".
>
> > Comments ?
>
> This mirrors my results, which is encouraging. The -mm2 patch seems to
> improve performance under write pressure quite a bit. I am attaching Con
> Kolivas' patch to 41-mm2 in case you missed it, as you can note from the
> results on the website, it improves things beyond -mm2. If you decide to
> run this version I'd like to see the result. I believe I had to use the
> "-l" patch option to ignore blank mismatches to get this to work, and I've
> cleaned up another mailing funny as well.
Ok, thanks for the patch.
I try it and I back with the result.
Ciao,
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
2002-10-13 11:57 Paolo Ciarrocchi
@ 2002-10-13 13:38 ` Bill Davidsen
0 siblings, 0 replies; 11+ messages in thread
From: Bill Davidsen @ 2002-10-13 13:38 UTC (permalink / raw)
To: Paolo Ciarrocchi; +Cc: linux-kernel
On Sun, 13 Oct 2002, Paolo Ciarrocchi wrote:
> Hi David,
thanks for the quick results, but it's Bill...
> I think your benchmark is very intersting.
> Here goes my results:
> It seems that 2.5.42-mm2 is the "winner".
> Comments ?
This mirrors my results, which is encouraging. The -mm2 patch seems to
improve performance under write pressure quite a bit. I am attaching Con
Kolivas' patch to 41-mm2 in case you missed it, as you can note from the
results on the website, it improves things beyond -mm2. If you decide to
run this version I'd like to see the result. I believe I had to use the
"-l" patch option to ignore blank mismatches to get this to work, and I've
cleaned up another mailing funny as well.
--- linux-2.5.41/mm/vmscan.c 2002-10-11 09:11:20.000000000 +1000
+++ linux-2.5.41-new/mm/vmscan.c 2002-10-11 00:51:06.000000000 +1000
@@ -44,7 +44,8 @@
/*
* From 0 .. 100. Higher means more swappy.
*/
-int vm_swappiness = 50;
+int vm_swappiness = 0;
+int vm_swap_feedback;
static long total_memory;
#ifdef ARCH_HAS_PREFETCH
@@ -587,7 +588,18 @@
* A 100% value of vm_swappiness will override this algorithm almost
* altogether.
*/
- swap_tendency = mapped_ratio / 2 + distress + vm_swappiness;
+ swap_tendency = mapped_ratio / 2 + distress;
+
+ vm_swap_feedback = (swap_tendency - 50)/10;
+ vm_swappiness += vm_swap_feedback;
+ if (vm_swappiness < 0){
+ vm_swappiness = 0;
+ }
+ else
+ if (vm_swappiness > 100){
+ vm_swappiness = 100;
+ }
+ swap_tendency += vm_swappiness;
/*
* Well that all made sense. Now for some magic numbers. Use the
--
bill davidsen <davidsen@tmr.com>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re:Benchmark results from resp1 trivial response time test
@ 2002-10-13 11:57 Paolo Ciarrocchi
2002-10-13 13:38 ` Bill Davidsen
0 siblings, 1 reply; 11+ messages in thread
From: Paolo Ciarrocchi @ 2002-10-13 11:57 UTC (permalink / raw)
To: linux-kernel; +Cc: davidsen
Hi David,
I think your benchmark is very intersting.
Here goes my results:
--- 2.4.19 ---
Starting 1 CPU run with 251 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 119.653 124.207 119.786 120.644 0.002 1.000
smallwrite 121.051 190.696 144.469 152.062 0.031 1.260
largewrite 117.412 11871.027 1113.563 3362.302 4.870 27.870
cpuload 119.691 313.702 178.513 190.324 0.081 1.578
spawnload 114.624 243.985 119.747 154.604 0.057 1.281
8ctx-mem 652.737 7564.493 1234.938 2527.273 2.904 20.948
2ctx-mem 126.914 8428.021 620.418 2941.043 3.890 24.378
--- 2.5.41 ---
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.607 114.030 113.774 113.783 0.000 1.000
smallwrite 117.882 526.275 121.038 208.868 0.178 1.836
largewrite 1674.216 88428.871 14794.752 27750.852 34.851 243.892
cpuload 104.225 158.400 104.478 115.269 0.024 1.013
spawnload 105.933 166.818 106.682 118.452 0.027 1.041
8ctx-mem 116.458 8893.645 120.670 1875.275 3.923 16.481
2ctx-mem 116.847 10174.152 121.309 2130.303 4.497 18.722
--- 2.5.42 ---
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.676 113.948 113.737 113.774 0.000 1.000
smallwrite 115.352 189.983 119.500 141.645 0.036 1.245
largewrite 19012.080 45528.170 30311.890 31472.632 9.646 276.625
cpuload 103.328 1267.116 103.971 336.462 0.520 2.957
spawnload 105.196 167.400 105.787 117.983 0.028 1.037
8ctx-mem 121.166 8616.232 126.489 1824.054 3.797 16.032
2ctx-mem 115.162 10291.470 119.560 2152.930 4.550 18.923
--- 2.5.42-mm2 ---
Starting 1 CPU run with 250 MB RAM, minimum 5 data points at 20 sec intervals
_____________ delay ms. ____________
Test low high median average S.D. ratio
noload 113.658 114.500 113.737 113.866 0.000 1.000
smallwrite 117.220 245.110 119.017 144.541 0.056 1.269
largewrite 114.841 181.420 118.746 131.568 0.028 1.155
cpuload 104.583 159.080 104.890 115.694 0.024 1.016
spawnload 108.790 166.141 109.655 120.698 0.025 1.060
8ctx-mem 114.430 8484.873 784.141 2066.476 3.604 18.148
2ctx-mem 121.388 8453.018 1414.443 2383.773 3.459 20.935
It seems that 2.5.42-mm2 is the "winner".
Comments ?
Paolo
--
Get your free email from www.linuxmail.org
Powered by Outblaze
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2002-10-17 20:11 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-10-13 16:47 Re:Benchmark results from resp1 trivial response time test Paolo Ciarrocchi
2002-10-13 21:02 ` Bill Davidsen
-- strict thread matches above, loose matches on Subject: below --
2002-10-17 20:16 Paolo Ciarrocchi
2002-10-15 21:03 Paolo Ciarrocchi
2002-10-15 19:15 Paolo Ciarrocchi
2002-10-14 21:30 Paolo Ciarrocchi
2002-10-15 1:31 ` Bill Davidsen
2002-10-14 7:38 Paolo Ciarrocchi
2002-10-13 15:25 Paolo Ciarrocchi
2002-10-13 11:57 Paolo Ciarrocchi
2002-10-13 13:38 ` Bill Davidsen
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).