We have an application that will write and read from every sector on a drive. The application can perform these tasks on multiple drives at the same time. It is designed to run on top of the Linux Kernel, which we periodically update so that we can get the latest device drivers. When performing the last update from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives decreased by some 40% (took 3 hours and 11 minutes to write and read from 5 drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able to determine that the issue was in the 2.6.37 Kernel as I was able to run it with the 2.6.36.4 kernel, and it had the better performance. After seeing that I/O throttling was introduced in the 2.6.37 Kernel, I naturally suspected that. However, by default, all the throttling was turned off (I attached the actual .config that was used to build the kernel). I then tried to turn on the throttling and set it to a high number to see what would happen. When I did that, I was able to reduce the time from 3 hours and 11 minutes to 2 hours and 50 minutes. There seems to be something there that changed that is impacting performance on multiple drives. When we do this same test with only one drive, the performance is identical between the systems. This issue still occurs on Kernel 3.0.2.
Created attachment 69622 [details] The .config used to build the 2.6.37 Kernel This is the .config used to build the 2.6.37 Kernel
Created attachment 69632 [details] The log of system where performance was good
Created attachment 69642 [details] The log of system where performance was degraded.
(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface). On Mon, 22 Aug 2011 15:20:41 GMT bugzilla-daemon@bugzilla.kernel.org wrote: > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > Summary: Performance of writing and reading from multiple > drives decreases by 40% when going from Linux Kernel > 2.6.36.4 to 2.6.37 (and beyond) > Product: IO/Storage > Version: 2.5 > Kernel Version: 2.6.37 > Platform: All > OS/Version: Linux > Tree: Mainline > Status: NEW > Severity: normal > Priority: P1 > Component: SCSI > AssignedTo: linux-scsi@vger.kernel.org > ReportedBy: mpete_06@hotmail.com > Regression: No > > > We have an application that will write and read from every sector on a drive. > The application can perform these tasks on multiple drives at the same time. > It is designed to run on top of the Linux Kernel, which we periodically > update > so that we can get the latest device drivers. When performing the last > update > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > decreased by some 40% (took 3 hours and 11 minutes to write and read from 5 > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able to > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > the 2.6.36.4 kernel, and it had the better performance. After seeing that > I/O > throttling was introduced in the 2.6.37 Kernel, I naturally suspected that. > However, by default, all the throttling was turned off (I attached the actual > .config that was used to build the kernel). I then tried to turn on the > throttling and set it to a high number to see what would happen. When I did > that, I was able to reduce the time from 3 hours and 11 minutes to 2 hours > and > 50 minutes. There seems to be something there that changed that is impacting > performance on multiple drives. When we do this same test with only one > drive, > the performance is identical between the systems. This issue still occurs on > Kernel 3.0.2. > Are you able to determine whether this regression is due to slower reading, to slower writing or to both? Thanks.
Reply-To: vgoyal@redhat.com On Mon, Aug 22, 2011 at 02:49:56PM -0500, Mark Petersen wrote: > > The majority of the slowdown we found is coming during the writing as we were > doing limited reading for the purpose of the testing. It may be that it > happens in both areas, but we did not do extensive testing with the reading > portion of it. What kind of writes these are? Write slowdown by 40%. Somehow now a days barriers/flush/fua comes to my mind. Any changes there w.r.t your setup? Recently Jeff moyer and Mike Snitzer had discovered and fixed a slowdown in a dm-multipath and disks not having write caches. I guess that's not your setup. Though mentioning it does not harm. Thanks Vivek > > > Date: Mon, 22 Aug 2011 12:24:43 -0700 > > From: akpm@linux-foundation.org > > To: mpete_06@hotmail.com > > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; vgoyal@redhat.com; > linux-mm@kvack.org; linux-scsi@vger.kernel.org > > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and > reading from multiple drives decreases by 40% when going from Linux Kernel > 2.6.36.4 to 2.6.37 (and beyond) > > > > > > (switched to email. Please respond via emailed reply-to-all, not via the > > bugzilla web interface). > > > > On Mon, 22 Aug 2011 15:20:41 GMT > > bugzilla-daemon@bugzilla.kernel.org wrote: > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > > > Summary: Performance of writing and reading from multiple > > > drives decreases by 40% when going from Linux Kernel > > > 2.6.36.4 to 2.6.37 (and beyond) > > > Product: IO/Storage > > > Version: 2.5 > > > Kernel Version: 2.6.37 > > > Platform: All > > > OS/Version: Linux > > > Tree: Mainline > > > Status: NEW > > > Severity: normal > > > Priority: P1 > > > Component: SCSI > > > AssignedTo: linux-scsi@vger.kernel.org > > > ReportedBy: mpete_06@hotmail.com > > > Regression: No > > > > > > > > > We have an application that will write and read from every sector on a > drive. > > > The application can perform these tasks on multiple drives at the same > time. > > > It is designed to run on top of the Linux Kernel, which we periodically > update > > > so that we can get the latest device drivers. When performing the last > update > > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > > > decreased by some 40% (took 3 hours and 11 minutes to write and read from > 5 > > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able > to > > > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > > > the 2.6.36.4 kernel, and it had the better performance. After seeing > that I/O > > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected > that. > > > However, by default, all the throttling was turned off (I attached the > actual > > > .config that was used to build the kernel). I then tried to turn on the > > > throttling and set it to a high number to see what would happen. When I > did > > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 > hours and > > > 50 minutes. There seems to be something there that changed that is > impacting > > > performance on multiple drives. When we do this same test with only one > drive, > > > the performance is identical between the systems. This issue still > occurs on > > > Kernel 3.0.2. > > > > > > > Are you able to determine whether this regression is due to slower > > reading, to slower writing or to both? > > > > Thanks. >
Reply-To: vgoyal@redhat.com On Mon, Aug 22, 2011 at 12:24:43PM -0700, Andrew Morton wrote: > > (switched to email. Please respond via emailed reply-to-all, not via the > bugzilla web interface). > > On Mon, 22 Aug 2011 15:20:41 GMT > bugzilla-daemon@bugzilla.kernel.org wrote: > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > Summary: Performance of writing and reading from multiple > > drives decreases by 40% when going from Linux Kernel > > 2.6.36.4 to 2.6.37 (and beyond) > > Product: IO/Storage > > Version: 2.5 > > Kernel Version: 2.6.37 > > Platform: All > > OS/Version: Linux > > Tree: Mainline > > Status: NEW > > Severity: normal > > Priority: P1 > > Component: SCSI > > AssignedTo: linux-scsi@vger.kernel.org > > ReportedBy: mpete_06@hotmail.com > > Regression: No > > > > > > We have an application that will write and read from every sector on a > drive. > > The application can perform these tasks on multiple drives at the same > time. > > It is designed to run on top of the Linux Kernel, which we periodically > update > > so that we can get the latest device drivers. When performing the last > update > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > > decreased by some 40% (took 3 hours and 11 minutes to write and read from 5 > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able to > > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > > the 2.6.36.4 kernel, and it had the better performance. After seeing that > I/O > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected that. > > However, by default, all the throttling was turned off (I attached the > actual > > .config that was used to build the kernel). I then tried to turn on the > > throttling and set it to a high number to see what would happen. When I > did > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 hours > and > > 50 minutes. There seems to be something there that changed that is > impacting > > performance on multiple drives. When we do this same test with only one > drive, > > the performance is identical between the systems. This issue still occurs > on > > Kernel 3.0.2. > > > > Are you able to determine whether this regression is due to slower > reading, to slower writing or to both? Mark, As your initial comment says that you see 40% regression even when block throttling infrastructure is not enabled, I think it is not related to throttling as blk_throtl_bio() is null when BLK_DEV_THROTTLING=n. What IO scheduler are you using? Can you try switching IO scheduler to deadline and see if regression is still there. Trying to figure out if it has anything to do with IO scheduler. What file system are you using with what options? Are you using device mapper to create some special configuration on multiple disks? Also can you take a trace (blktrace) of any of the disks for 30 seconds both without regression and after regression and upload it somewhere. Staring at it might give some clues. Thanks Vivek
The majority of the slowdown we found is coming during the writing as we were doing limited reading for the purpose of the testing. It may be that it happens in both areas, but we did not do extensive testing with the reading portion of it. > Date: Mon, 22 Aug 2011 12:24:43 -0700 > From: akpm@linux-foundation.org > To: mpete_06@hotmail.com > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; vgoyal@redhat.com; > linux-mm@kvack.org; linux-scsi@vger.kernel.org > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and reading > from multiple drives decreases by 40% when going from Linux Kernel 2.6.36.4 > to 2.6.37 (and beyond) > > > (switched to email. Please respond via emailed reply-to-all, not via the > bugzilla web interface). > > On Mon, 22 Aug 2011 15:20:41 GMT > bugzilla-daemon@bugzilla.kernel.org wrote: > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > Summary: Performance of writing and reading from multiple > > drives decreases by 40% when going from Linux Kernel > > 2.6.36.4 to 2.6.37 (and beyond) > > Product: IO/Storage > > Version: 2.5 > > Kernel Version: 2.6.37 > > Platform: All > > OS/Version: Linux > > Tree: Mainline > > Status: NEW > > Severity: normal > > Priority: P1 > > Component: SCSI > > AssignedTo: linux-scsi@vger.kernel.org > > ReportedBy: mpete_06@hotmail.com > > Regression: No > > > > > > We have an application that will write and read from every sector on a > drive. > > The application can perform these tasks on multiple drives at the same > time. > > It is designed to run on top of the Linux Kernel, which we periodically > update > > so that we can get the latest device drivers. When performing the last > update > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > > decreased by some 40% (took 3 hours and 11 minutes to write and read from 5 > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able to > > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > > the 2.6.36.4 kernel, and it had the better performance. After seeing that > I/O > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected that. > > However, by default, all the throttling was turned off (I attached the > actual > > .config that was used to build the kernel). I then tried to turn on the > > throttling and set it to a high number to see what would happen. When I > did > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 hours > and > > 50 minutes. There seems to be something there that changed that is > impacting > > performance on multiple drives. When we do this same test with only one > drive, > > the performance is identical between the systems. This issue still occurs > on > > Kernel 3.0.2. > > > > Are you able to determine whether this regression is due to slower > reading, to slower writing or to both? > > Thanks.
The writes we are performing are SCSI commands directly to the device, one sector at a time. The only thing we changed between our updates was the Kernel itself, which we leave everything in there at its default value if it is enabled (we disable a great many things we don't need). The latest version I tried that still showed the issue was v3.0.1. Thanks, Mark > Date: Mon, 22 Aug 2011 15:56:51 -0400 > From: vgoyal@redhat.com > To: mpete_06@hotmail.com > CC: akpm@linux-foundation.org; bugme-daemon@bugzilla.kernel.org; > axboe@kernel.dk; linux-mm@kvack.org; linux-scsi@vger.kernel.org > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and reading > from multiple drives decreases by 40% when going from Linux Kernel 2.6.36.4 > to 2.6.37 (and beyond) > > On Mon, Aug 22, 2011 at 02:49:56PM -0500, Mark Petersen wrote: > > > > The majority of the slowdown we found is coming during the writing as we > were doing limited reading for the purpose of the testing. It may be that it > happens in both areas, but we did not do extensive testing with the reading > portion of it. > > What kind of writes these are? Write slowdown by 40%. Somehow now a days > barriers/flush/fua comes to my mind. Any changes there w.r.t your setup? > > Recently Jeff moyer and Mike Snitzer had discovered and fixed a slowdown > in a dm-multipath and disks not having write caches. I guess that's not > your setup. Though mentioning it does not harm. > > Thanks > Vivek > > > > > > > Date: Mon, 22 Aug 2011 12:24:43 -0700 > > > From: akpm@linux-foundation.org > > > To: mpete_06@hotmail.com > > > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; vgoyal@redhat.com; > linux-mm@kvack.org; linux-scsi@vger.kernel.org > > > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and > reading from multiple drives decreases by 40% when going from Linux Kernel > 2.6.36.4 to 2.6.37 (and beyond) > > > > > > > > > (switched to email. Please respond via emailed reply-to-all, not via the > > > bugzilla web interface). > > > > > > On Mon, 22 Aug 2011 15:20:41 GMT > > > bugzilla-daemon@bugzilla.kernel.org wrote: > > > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > > > > > Summary: Performance of writing and reading from multiple > > > > drives decreases by 40% when going from Linux > Kernel > > > > 2.6.36.4 to 2.6.37 (and beyond) > > > > Product: IO/Storage > > > > Version: 2.5 > > > > Kernel Version: 2.6.37 > > > > Platform: All > > > > OS/Version: Linux > > > > Tree: Mainline > > > > Status: NEW > > > > Severity: normal > > > > Priority: P1 > > > > Component: SCSI > > > > AssignedTo: linux-scsi@vger.kernel.org > > > > ReportedBy: mpete_06@hotmail.com > > > > Regression: No > > > > > > > > > > > > We have an application that will write and read from every sector on a > drive. > > > > The application can perform these tasks on multiple drives at the same > time. > > > > It is designed to run on top of the Linux Kernel, which we periodically > update > > > > so that we can get the latest device drivers. When performing the last > update > > > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of > drives > > > > decreased by some 40% (took 3 hours and 11 minutes to write and read > from 5 > > > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was > able to > > > > determine that the issue was in the 2.6.37 Kernel as I was able to run > it with > > > > the 2.6.36.4 kernel, and it had the better performance. After seeing > that I/O > > > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected > that. > > > > However, by default, all the throttling was turned off (I attached the > actual > > > > .config that was used to build the kernel). I then tried to turn on > the > > > > throttling and set it to a high number to see what would happen. When > I did > > > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 > hours and > > > > 50 minutes. There seems to be something there that changed that is > impacting > > > > performance on multiple drives. When we do this same test with only > one drive, > > > > the performance is identical between the systems. This issue still > occurs on > > > > Kernel 3.0.2. > > > > > > > > > > Are you able to determine whether this regression is due to slower > > > reading, to slower writing or to both? > > > > > > Thanks. > >
We were using the default CFQ scheduler. I will change it to deadline and see what happens. Also, we are not using a file system to perform the writes, rather we are sending SCSI commands directly to the devices, nor are we doing anything special to the disks with a device mapper. We simply write to each one on a different thread one sector at a time. I will attempt to get the trace and will add it if I can. Thanks, Mark > Date: Mon, 22 Aug 2011 15:48:54 -0400 > From: vgoyal@redhat.com > To: mpete_06@hotmail.com > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; linux-mm@kvack.org; > linux-scsi@vger.kernel.org; akpm@linux-foundation.org > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and reading > from multiple drives decreases by 40% when going from Linux Kernel 2.6.36.4 > to 2.6.37 (and beyond) > > On Mon, Aug 22, 2011 at 12:24:43PM -0700, Andrew Morton wrote: > > > > (switched to email. Please respond via emailed reply-to-all, not via the > > bugzilla web interface). > > > > On Mon, 22 Aug 2011 15:20:41 GMT > > bugzilla-daemon@bugzilla.kernel.org wrote: > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > > > Summary: Performance of writing and reading from multiple > > > drives decreases by 40% when going from Linux Kernel > > > 2.6.36.4 to 2.6.37 (and beyond) > > > Product: IO/Storage > > > Version: 2.5 > > > Kernel Version: 2.6.37 > > > Platform: All > > > OS/Version: Linux > > > Tree: Mainline > > > Status: NEW > > > Severity: normal > > > Priority: P1 > > > Component: SCSI > > > AssignedTo: linux-scsi@vger.kernel.org > > > ReportedBy: mpete_06@hotmail.com > > > Regression: No > > > > > > > > > We have an application that will write and read from every sector on a > drive. > > > The application can perform these tasks on multiple drives at the same > time. > > > It is designed to run on top of the Linux Kernel, which we periodically > update > > > so that we can get the latest device drivers. When performing the last > update > > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > > > decreased by some 40% (took 3 hours and 11 minutes to write and read from > 5 > > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able > to > > > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > > > the 2.6.36.4 kernel, and it had the better performance. After seeing > that I/O > > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected > that. > > > However, by default, all the throttling was turned off (I attached the > actual > > > .config that was used to build the kernel). I then tried to turn on the > > > throttling and set it to a high number to see what would happen. When I > did > > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 > hours and > > > 50 minutes. There seems to be something there that changed that is > impacting > > > performance on multiple drives. When we do this same test with only one > drive, > > > the performance is identical between the systems. This issue still > occurs on > > > Kernel 3.0.2. > > > > > > > Are you able to determine whether this regression is due to slower > > reading, to slower writing or to both? > > Mark, > > As your initial comment says that you see 40% regression even when block > throttling infrastructure is not enabled, I think it is not related to > throttling as blk_throtl_bio() is null when BLK_DEV_THROTTLING=n. > > What IO scheduler are you using? Can you try switching IO scheduler to > deadline and see if regression is still there. Trying to figure out if > it has anything to do with IO scheduler. > > What file system are you using with what options? Are you using device > mapper to create some special configuration on multiple disks? > > Also can you take a trace (blktrace) of any of the disks for 30 seconds > both without regression and after regression and upload it somewhere. > Staring at it might give some clues. > > Thanks > Vivek
Reply-To: James.Bottomley@HansenPartnership.com On Mon, 2011-08-22 at 21:07 +0000, bugzilla-daemon@bugzilla.kernel.org wrote: > We were using the default CFQ scheduler. I will change it to deadline and > see > what happens. Also, we are not using a file system to perform the writes, > rather we are sending SCSI commands directly to the devices, nor are we doing > anything special to the disks with a device mapper. We simply write to each > one on a different thread one sector at a time. > > I will attempt to get the trace and will add it if I can. If you're just using a SG_IO type scsi command, then virtually nothing we've done in block/scsi should affect this; certainly not the elevators since we don't do merging on SG_IO, we just use them for queueing. Since you say "write to each one on a different thread", it's possible that some scheduler change may have impacted this. James
I was finally able to run it with the deadline scheduler, and got the same performance. Unfortunately, I am not able to use the blktrace tool as it requires a version of libc that we do not have on the system (we have 2.5 and it requires at least 2.7). Is there anything else I can use to trace it? Thanks, Mark > Date: Mon, 22 Aug 2011 15:48:54 -0400 > From: vgoyal@redhat.com > To: mpete_06@hotmail.com > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; linux-mm@kvack.org; > linux-scsi@vger.kernel.org; akpm@linux-foundation.org > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and reading > from multiple drives decreases by 40% when going from Linux Kernel 2.6.36.4 > to 2.6.37 (and beyond) > > On Mon, Aug 22, 2011 at 12:24:43PM -0700, Andrew Morton wrote: > > > > (switched to email. Please respond via emailed reply-to-all, not via the > > bugzilla web interface). > > > > On Mon, 22 Aug 2011 15:20:41 GMT > > bugzilla-daemon@bugzilla.kernel.org wrote: > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > > > Summary: Performance of writing and reading from multiple > > > drives decreases by 40% when going from Linux Kernel > > > 2.6.36.4 to 2.6.37 (and beyond) > > > Product: IO/Storage > > > Version: 2.5 > > > Kernel Version: 2.6.37 > > > Platform: All > > > OS/Version: Linux > > > Tree: Mainline > > > Status: NEW > > > Severity: normal > > > Priority: P1 > > > Component: SCSI > > > AssignedTo: linux-scsi@vger.kernel.org > > > ReportedBy: mpete_06@hotmail.com > > > Regression: No > > > > > > > > > We have an application that will write and read from every sector on a > drive. > > > The application can perform these tasks on multiple drives at the same > time. > > > It is designed to run on top of the Linux Kernel, which we periodically > update > > > so that we can get the latest device drivers. When performing the last > update > > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of drives > > > decreased by some 40% (took 3 hours and 11 minutes to write and read from > 5 > > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was able > to > > > determine that the issue was in the 2.6.37 Kernel as I was able to run it > with > > > the 2.6.36.4 kernel, and it had the better performance. After seeing > that I/O > > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected > that. > > > However, by default, all the throttling was turned off (I attached the > actual > > > .config that was used to build the kernel). I then tried to turn on the > > > throttling and set it to a high number to see what would happen. When I > did > > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 > hours and > > > 50 minutes. There seems to be something there that changed that is > impacting > > > performance on multiple drives. When we do this same test with only one > drive, > > > the performance is identical between the systems. This issue still > occurs on > > > Kernel 3.0.2. > > > > > > > Are you able to determine whether this regression is due to slower > > reading, to slower writing or to both? > > Mark, > > As your initial comment says that you see 40% regression even when block > throttling infrastructure is not enabled, I think it is not related to > throttling as blk_throtl_bio() is null when BLK_DEV_THROTTLING=n. > > What IO scheduler are you using? Can you try switching IO scheduler to > deadline and see if regression is still there. Trying to figure out if > it has anything to do with IO scheduler. > > What file system are you using with what options? Are you using device > mapper to create some special configuration on multiple disks? > > Also can you take a trace (blktrace) of any of the disks for 30 seconds > both without regression and after regression and upload it somewhere. > Staring at it might give some clues. > > Thanks > Vivek
Reply-To: vgoyal@redhat.com On Wed, Aug 24, 2011 at 03:11:57PM -0500, Mark Petersen wrote: > > I was finally able to run it with the deadline scheduler, and got the same > performance. You mean you see 40% regression even with deadline? If yes, then it is not a IO scheduler specific issue. > Unfortunately, I am not able to use the blktrace tool as it requires a > version of libc that we do not have on the system (we have 2.5 and it > requires at least 2.7). Is there anything else I can use to trace it? > You can try using tracing functionality. - mount -t debugfs none /sys/kernel/debug - Enable tracing on the disk you are doing IO to. echo 1 > /sys/block/sda/trace/enable - Enable block traces echo blk > /sys/kernel/debug/tracing/current_tracer - cat /sys/kernel/debug/tracing/trace_pipe > /tmp/trace_output Let it run for few seconds. Interrupt and kill cat process. /tmp/trace_output should have useful tracing info. Thanks Vivek > Thanks, > Mark > > > Date: Mon, 22 Aug 2011 15:48:54 -0400 > > From: vgoyal@redhat.com > > To: mpete_06@hotmail.com > > CC: bugme-daemon@bugzilla.kernel.org; axboe@kernel.dk; linux-mm@kvack.org; > linux-scsi@vger.kernel.org; akpm@linux-foundation.org > > Subject: Re: [Bugme-new] [Bug 41552] New: Performance of writing and > reading from multiple drives decreases by 40% when going from Linux Kernel > 2.6.36.4 to 2.6.37 (and beyond) > > > > On Mon, Aug 22, 2011 at 12:24:43PM -0700, Andrew Morton wrote: > > > > > > (switched to email. Please respond via emailed reply-to-all, not via the > > > bugzilla web interface). > > > > > > On Mon, 22 Aug 2011 15:20:41 GMT > > > bugzilla-daemon@bugzilla.kernel.org wrote: > > > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=41552 > > > > > > > > Summary: Performance of writing and reading from multiple > > > > drives decreases by 40% when going from Linux > Kernel > > > > 2.6.36.4 to 2.6.37 (and beyond) > > > > Product: IO/Storage > > > > Version: 2.5 > > > > Kernel Version: 2.6.37 > > > > Platform: All > > > > OS/Version: Linux > > > > Tree: Mainline > > > > Status: NEW > > > > Severity: normal > > > > Priority: P1 > > > > Component: SCSI > > > > AssignedTo: linux-scsi@vger.kernel.org > > > > ReportedBy: mpete_06@hotmail.com > > > > Regression: No > > > > > > > > > > > > We have an application that will write and read from every sector on a > drive. > > > > The application can perform these tasks on multiple drives at the same > time. > > > > It is designed to run on top of the Linux Kernel, which we periodically > update > > > > so that we can get the latest device drivers. When performing the last > update > > > > from 2.6.33.2 to 2.6.37, we found that the performance of a set of > drives > > > > decreased by some 40% (took 3 hours and 11 minutes to write and read > from 5 > > > > drives on 2.6.37 versus 2 hours and 12 minutes on 2.6.33.3). I was > able to > > > > determine that the issue was in the 2.6.37 Kernel as I was able to run > it with > > > > the 2.6.36.4 kernel, and it had the better performance. After seeing > that I/O > > > > throttling was introduced in the 2.6.37 Kernel, I naturally suspected > that. > > > > However, by default, all the throttling was turned off (I attached the > actual > > > > .config that was used to build the kernel). I then tried to turn on > the > > > > throttling and set it to a high number to see what would happen. When > I did > > > > that, I was able to reduce the time from 3 hours and 11 minutes to 2 > hours and > > > > 50 minutes. There seems to be something there that changed that is > impacting > > > > performance on multiple drives. When we do this same test with only > one drive, > > > > the performance is identical between the systems. This issue still > occurs on > > > > Kernel 3.0.2. > > > > > > > > > > Are you able to determine whether this regression is due to slower > > > reading, to slower writing or to both? > > > > Mark, > > > > As your initial comment says that you see 40% regression even when block > > throttling infrastructure is not enabled, I think it is not related to > > throttling as blk_throtl_bio() is null when BLK_DEV_THROTTLING=n. > > > > What IO scheduler are you using? Can you try switching IO scheduler to > > deadline and see if regression is still there. Trying to figure out if > > it has anything to do with IO scheduler. > > > > What file system are you using with what options? Are you using device > > mapper to create some special configuration on multiple disks? > > > > Also can you take a trace (blktrace) of any of the disks for 30 seconds > > both without regression and after regression and upload it somewhere. > > Staring at it might give some clues. > > > > Thanks > > Vivek >