From: Johannes Thumshirn <jthumshirn@suse.de>
To: dm-devel@redhat.com
Cc: Mike Snitzer <snitzer@redhat.com>, Hannes Reinecke <hare@suse.de>,
linux-scsi@vger.kernel.org, sagig@dev.mellanox.co.il,
jmoyer@redhat.com, linux-block@vger.kernel.org,
j-nomura@ce.jp.nec.com
Subject: Re: [dm-devel] [RFC PATCH 0/4] dm mpath: vastly improve blk-mq IO performance
Date: Fri, 08 Apr 2016 13:42:16 +0200 [thread overview]
Message-ID: <2760420.6GyDefUaH5@c203> (raw)
In-Reply-To: <20160407153448.GA30510@redhat.com>
Ladies and Gentlemen,
To show off some numbers from our testing:
All tests are performed against the cache of the Array, not the disks as we
wanted to test the Linux stack not the Disk Array.
All single queue tests have been performed with the deadline I/O Scheduler.
Comments welcome, have fun reading :-)
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM MQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 785136 | 3066.1MB/s |
8k | 734983 | 5742.6MB/s |
16k | 398516 | 6226.9MB/s |
32k | 200589 | 6268.5MB/s |
64k | 100417 | 6276.2MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 788620 | 3080.6MB/s |
8k | 736359 | 5752.9MB/s |
16k | 398597 | 6228.8MB/s |
32k | 200487| 6265.3MB/s |
64k | 100402 | 6275.2MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 242105 | 968423KB/s |
8k | 222290 | 1736.7MB/s |
16k | 178191 | 2784.3MB/s |
32k | 133619 | 4175.7MB/s |
64k | 97693 | 6105.9MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134788 | 539155KB/s |
8k | 132361 | 1034.8MB/s |
16k | 129941 | 2030.4MB/s |
32k | 128288 | 4009.4MB/s |
64k | 97776 | 6111.0MB/s |
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM SQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 112402 | 449608KB/s |
8k | 112818 | 902551KB/s |
16k | 111885 | 1748.3MB/s |
32k | 188015 | 5875.6MB/s |
64k | 99021 | 6188.9MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 115046 | 460186KB/s |
8k | 113974 | 911799KB/s |
16k | 113374 | 1771.5MB/s |
32k | 192932 | 6029.2MB/s |
64k | 100474 | 6279.7MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 114284 | 457138KB/s |
8k | 113992 | 911944KB/s |
16k | 113715 | 1776.9MB/s |
32k | 130402 | 4075.9MB/s |
64k | 92243 | 5765.3MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 115540 | 462162KB/s |
8k | 114243 | 913951KB/s |
16k | 300153 | 4689.1MB/s |
32k | 141069 | 4408.5MB/s |
64k | 97620 | 6101.3MB/s |
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM MQ previous patchset:
============================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 782733 | 3057.6MB/s |
8k | 732143 | 5719.9MB/s |
16k | 398314 | 6223.7MB/s |
32k | 200538 | 6266.9MB/s |
64k | 100422 | 6276.5MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 786707 | 3073.8MB/s |
8k | 730579 | 5707.7MB/s |
16k | 398799 | 6231.3MB/s |
32k | 200518 | 6266.2MB/s |
64k | 100397 | 6274.9MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 242426 | 969707KB/s |
8k | 223079 | 1742.9MB/s |
16k | 177889 | 2779.6MB/s |
32k | 133637 | 4176.2MB/s |
64k | 97727 | 6107.1MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134360 | 537442KB/s |
8k | 129738 | 1013.6MB/s |
16k | 129746 | 2027.3MB/s |
32k | 127875 | 3996.1MB/s |
64k | 97683 | 6105.3MB/s |
Emulex 16GBit FC HBA NetApp EF560 Array w/ DM MQ this patchset :
========================================================
[Beware, this is with Hannes' lockless lpfc patches, which are not upstream as
they're quite experimental, but are good at showing the capability of the new
dm-mpath]
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 939752 | 3670.1MB/s |
8k | 741462 | 5792.7MB/s |
16k | 399285 | 6238.9MB/s |
32k | 196490 | 6140.4MB/s |
64k | 100325 | 6270.4MB/s |
Sequential Read:
BS | IOPS | BW
------+------------+-------------------+
4k | 926222 | 3618.6MB/s |
8k | 750125 | 5860.4MB/s |
16k | 397770 | 6215.2MB/s |
32k | 200130 | 6254.8MB/s |
64k | 100397 | 6274.9MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 251938 | 984.14MB/s |
8k | 226712 | 1771.2MB/s |
16k | 180739 | 2824.5MB/s |
32k | 133316 | 4166.2MB/s |
64k | 98738 | 6171.2MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134660 | 538643KB/s |
8k | 131585 | 1028.9MB/s |
16k | 131030 | 2047.4MB/s |
32k | 126987 | 3968.4MB/s |
64k | 98882 | 6180.2MB/s |
Emulex 16GBit FC HBA NetApp EF560 Array w/ DM SQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 101860 | 407443KB/s |
8k | 100242 | 801940KB/s |
16k | 99774 | 1558.1MB/s |
32k | 134304 | 4197.8MB/s |
64k | 100126 | 6257.1MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 285585 | 1115.6MB/s |
8k | 313619 | 2450.2MB/s |
16k | 190696 | 2979.7MB/s |
32k | 200569 | 6267.9MB/s |
64k | 100366 | 6272.1MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 98434 | 393738KB/s |
8k | 97286 | 778294KB/s |
16k | 97623 | 1525.4MB/s |
32k | 126999 | 3968.8MB/s |
64k | 94309 | 5894.4MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 330066 | 1289.4MB/s |
8k | 452792 | 3537.5MB/s |
16k | 334481 | 5226.3MB/s |
32k | 186916 | 5841.2MB/s |
64k | 88698 | 5543.7MB/s |
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N�rnberg
GF: Felix Imend�rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N�rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
WARNING: multiple messages have this Message-ID (diff)
From: Johannes Thumshirn <jthumshirn@suse.de>
To: dm-devel@redhat.com
Cc: Mike Snitzer <snitzer@redhat.com>, Hannes Reinecke <hare@suse.de>,
linux-scsi@vger.kernel.org, sagig@dev.mellanox.co.il,
jmoyer@redhat.com, linux-block@vger.kernel.org,
j-nomura@ce.jp.nec.com
Subject: Re: [dm-devel] [RFC PATCH 0/4] dm mpath: vastly improve blk-mq IO performance
Date: Fri, 08 Apr 2016 13:42:16 +0200 [thread overview]
Message-ID: <2760420.6GyDefUaH5@c203> (raw)
In-Reply-To: <20160407153448.GA30510@redhat.com>
Ladies and Gentlemen,
To show off some numbers from our testing:
All tests are performed against the cache of the Array, not the disks as we
wanted to test the Linux stack not the Disk Array.
All single queue tests have been performed with the deadline I/O Scheduler.
Comments welcome, have fun reading :-)
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM MQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 785136 | 3066.1MB/s |
8k | 734983 | 5742.6MB/s |
16k | 398516 | 6226.9MB/s |
32k | 200589 | 6268.5MB/s |
64k | 100417 | 6276.2MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 788620 | 3080.6MB/s |
8k | 736359 | 5752.9MB/s |
16k | 398597 | 6228.8MB/s |
32k | 200487| 6265.3MB/s |
64k | 100402 | 6275.2MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 242105 | 968423KB/s |
8k | 222290 | 1736.7MB/s |
16k | 178191 | 2784.3MB/s |
32k | 133619 | 4175.7MB/s |
64k | 97693 | 6105.9MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134788 | 539155KB/s |
8k | 132361 | 1034.8MB/s |
16k | 129941 | 2030.4MB/s |
32k | 128288 | 4009.4MB/s |
64k | 97776 | 6111.0MB/s |
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM SQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 112402 | 449608KB/s |
8k | 112818 | 902551KB/s |
16k | 111885 | 1748.3MB/s |
32k | 188015 | 5875.6MB/s |
64k | 99021 | 6188.9MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 115046 | 460186KB/s |
8k | 113974 | 911799KB/s |
16k | 113374 | 1771.5MB/s |
32k | 192932 | 6029.2MB/s |
64k | 100474 | 6279.7MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 114284 | 457138KB/s |
8k | 113992 | 911944KB/s |
16k | 113715 | 1776.9MB/s |
32k | 130402 | 4075.9MB/s |
64k | 92243 | 5765.3MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 115540 | 462162KB/s |
8k | 114243 | 913951KB/s |
16k | 300153 | 4689.1MB/s |
32k | 141069 | 4408.5MB/s |
64k | 97620 | 6101.3MB/s |
QLogic 32GBit FC HBA NetApp EF560 Array w/ DM MQ previous patchset:
============================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 782733 | 3057.6MB/s |
8k | 732143 | 5719.9MB/s |
16k | 398314 | 6223.7MB/s |
32k | 200538 | 6266.9MB/s |
64k | 100422 | 6276.5MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 786707 | 3073.8MB/s |
8k | 730579 | 5707.7MB/s |
16k | 398799 | 6231.3MB/s |
32k | 200518 | 6266.2MB/s |
64k | 100397 | 6274.9MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 242426 | 969707KB/s |
8k | 223079 | 1742.9MB/s |
16k | 177889 | 2779.6MB/s |
32k | 133637 | 4176.2MB/s |
64k | 97727 | 6107.1MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134360 | 537442KB/s |
8k | 129738 | 1013.6MB/s |
16k | 129746 | 2027.3MB/s |
32k | 127875 | 3996.1MB/s |
64k | 97683 | 6105.3MB/s |
Emulex 16GBit FC HBA NetApp EF560 Array w/ DM MQ this patchset :
========================================================
[Beware, this is with Hannes' lockless lpfc patches, which are not upstream as
they're quite experimental, but are good at showing the capability of the new
dm-mpath]
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 939752 | 3670.1MB/s |
8k | 741462 | 5792.7MB/s |
16k | 399285 | 6238.9MB/s |
32k | 196490 | 6140.4MB/s |
64k | 100325 | 6270.4MB/s |
Sequential Read:
BS | IOPS | BW
------+------------+-------------------+
4k | 926222 | 3618.6MB/s |
8k | 750125 | 5860.4MB/s |
16k | 397770 | 6215.2MB/s |
32k | 200130 | 6254.8MB/s |
64k | 100397 | 6274.9MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 251938 | 984.14MB/s |
8k | 226712 | 1771.2MB/s |
16k | 180739 | 2824.5MB/s |
32k | 133316 | 4166.2MB/s |
64k | 98738 | 6171.2MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 134660 | 538643KB/s |
8k | 131585 | 1028.9MB/s |
16k | 131030 | 2047.4MB/s |
32k | 126987 | 3968.4MB/s |
64k | 98882 | 6180.2MB/s |
Emulex 16GBit FC HBA NetApp EF560 Array w/ DM SQ this patchset:
========================================================
Random Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 101860 | 407443KB/s |
8k | 100242 | 801940KB/s |
16k | 99774 | 1558.1MB/s |
32k | 134304 | 4197.8MB/s |
64k | 100126 | 6257.1MB/s |
Sequential Read:
BS | IOPS | BW |
------+------------+-------------------+
4k | 285585 | 1115.6MB/s |
8k | 313619 | 2450.2MB/s |
16k | 190696 | 2979.7MB/s |
32k | 200569 | 6267.9MB/s |
64k | 100366 | 6272.1MB/s |
Random Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 98434 | 393738KB/s |
8k | 97286 | 778294KB/s |
16k | 97623 | 1525.4MB/s |
32k | 126999 | 3968.8MB/s |
64k | 94309 | 5894.4MB/s |
Sequential Write:
BS | IOPS | BW |
------+------------+-------------------+
4k | 330066 | 1289.4MB/s |
8k | 452792 | 3537.5MB/s |
16k | 334481 | 5226.3MB/s |
32k | 186916 | 5841.2MB/s |
64k | 88698 | 5543.7MB/s |
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2016-04-08 11:42 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-31 20:04 [RFC PATCH 0/4] dm mpath: vastly improve blk-mq IO performance Mike Snitzer
2016-03-31 20:04 ` [RFC PATCH 1/4] dm mpath: switch to using bitops for state flags Mike Snitzer
2016-04-01 8:46 ` [dm-devel] " Johannes Thumshirn
2016-04-07 14:59 ` Hannes Reinecke
2016-03-31 20:04 ` [RFC PATCH 2/4] dm mpath: use atomic_t for counting members of 'struct multipath' Mike Snitzer
2016-04-01 8:48 ` [dm-devel] " Johannes Thumshirn
2016-04-07 15:02 ` Hannes Reinecke
2016-03-31 20:04 ` [RFC PATCH 3/4] dm mpath: move trigger_event member to the end " Mike Snitzer
2016-04-01 8:50 ` [dm-devel] " Johannes Thumshirn
2016-04-07 15:03 ` Hannes Reinecke
2016-03-31 20:04 ` [RFC PATCH 4/4] dm mpath: eliminate use of spinlock in IO fast-paths Mike Snitzer
2016-04-01 9:02 ` [dm-devel] " Johannes Thumshirn
2016-04-07 15:03 ` Hannes Reinecke
2016-04-01 8:12 ` [dm-devel] [RFC PATCH 0/4] dm mpath: vastly improve blk-mq IO performance Johannes Thumshirn
2016-04-01 13:22 ` Mike Snitzer
2016-04-01 13:37 ` Johannes Thumshirn
2016-04-01 14:14 ` Mike Snitzer
2016-04-07 14:58 ` Hannes Reinecke
2016-04-07 14:58 ` Hannes Reinecke
2016-04-07 15:34 ` Mike Snitzer
2016-04-08 11:42 ` Johannes Thumshirn [this message]
2016-04-08 11:42 ` [dm-devel] " Johannes Thumshirn
2016-04-08 19:29 ` Mike Snitzer
2016-04-13 7:03 ` [dm-devel] " Johannes Thumshirn
2016-04-13 7:03 ` Johannes Thumshirn
2016-05-09 15:48 ` Bart Van Assche
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2760420.6GyDefUaH5@c203 \
--to=jthumshirn@suse.de \
--cc=dm-devel@redhat.com \
--cc=hare@suse.de \
--cc=j-nomura@ce.jp.nec.com \
--cc=jmoyer@redhat.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=sagig@dev.mellanox.co.il \
--cc=snitzer@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.