This is a multi-part message in MIME format. --------------64D829377F18A02EC6553431 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit On 10/02/2017 11:05 AM, Jason Keltz wrote:

On 10/02/2017 11:00 AM, Yaniv Kaul wrote:

...

On Mon, Oct 2, 2017 at 5:57 PM, Jason Keltz <jas@cse.yorku.ca <mailto:jas@cse.yorku.ca>> wrote:

On 10/02/2017 10:51 AM, Yaniv Kaul wrote:

...

On Mon, Oct 2, 2017 at 5:14 PM, Jason Keltz <jas@cse.yorku.ca <mailto:jas@cse.yorku.ca>> wrote:

On 10/02/2017 01:22 AM, Yaniv Kaul wrote:

...

On Mon, Oct 2, 2017 at 5:11 AM, Jason Keltz <jas@cse.yorku.ca <mailto:jas@cse.yorku.ca>> wrote:

Hi.

For my data domain, I have one NFS server with a large RAID filesystem (9 TB). I'm only using 2 TB of that at the moment. Today, my NFS server hung with the following error:

xfs: possible memory allocation deadlock in kmem_alloc

Can you share more of the log so we'll see what happened before and after? Y.

Here is engine-log from yesterday.. the problem started around 14:29 PM. http://www.eecs.yorku.ca/~jas/ovirt-debug/10012017/engine-log.txt <http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/10012017/engine-log.txt>

Here is the vdsm log on one of the virtualization hosts, virt01: http://www.eecs.yorku.ca/~jas/ovirt-debug/10012017/vdsm.log.2 <http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/10012017/vdsm.log.2>

Doing further investigation, I found that the XFS error messages didn't start yesterday. You'll see they started at the very end of the day on September 23. See:

http://www.eecs.yorku.ca/~jas/ovirt-debug/messages-20170924 <http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/messages-20170924>

Our storage guys do NOT think it's an XFS fragmentation issue, but we'll be looking at it.

This is an interesting thread to read because the problem sounds quite similar:

http://oss.sgi.com/archives/xfs/2016-03/msg00447.html In particular, quoted from that:

XFS maintains the full extent list for an active inode in memory, As it is, yes, the memory allocation problem is with the in-core extent tree, and we've known about it for some time. The issue is that as memory gets fragmented, the top level indirection array grows too large to be allocated as a contiguous chunk. When this happens really depends on memory load, uptime and the way the extent tree is being modified.

So in my case, I have a bunch of big XFS disk images for virtual disks. As the files are big with many extents, keeping all that information in memory at the same time may be the culprit. Having many extents per se isn't the problem, but having enough memory to be able to store all the information simultaneously may be. Possible solutions would be to increase the default extent size of the volume (which I'm not sure how to do), defragment the disk, and hence less extents, or potentially add more memory to the file server. It has 64G.

...

...

Hmmm... almost sorry to hear that because that would be easy to "fix"...

...

They continued on the 24th, then on the 26th... I think there were a few "hangs" on those times that people were complaining about, but we didn't catch the problem. However, the errors hit big time yesterday at 14:27 PM... see here:

http://www.eecs.yorku.ca/~jas/ovirt-debug/messages-20171001 <http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/messages-20171001>

If you want any other logs, I'm happy to provide them. I just don't know exactly what to provide.

Do you know if I can run the XFS defrag command live? Rather than on a disk by disk, I'd rather just do it on the whole filesystem. There really aren't that many files since it's just ovirt disk images. However, I don't understand the implications to running VMs. I wouldn't want to do anything to create more downtime.

Should be enough to copy the disks to make them less fragmented.

Yes, but this requires downtime.. but there's plenty of additional storage, so this would fix things well.

Live storage migration could be used. Y.

...

I had upgraded the engine server + 4 virtualization hosts from 4.1.1 to current on September 20 along with upgrading them from CentOS 7.3 to CentOS 7.4. virtfs, the NFS file server, was running CentOS 7.3 and kernel vmlinuz-3.10.0-514.16.1.el7.x86_64. Only yesterday, did I upgrade it to CentOS 7.4 and hence kernel vmlinuz-3.10.0-693.2.2.el7.x86_64.

I believe the problem is fully XFS related, and not ovirt at all. Although, I must admit, ovirt didn't help either. When I rebooted the file server, the iso and export domains were immediately active, but the data domain took quite a long time. I kept trying to activate it, and it couldn't do it. I couldn't make a host an SPM. I found that the data domain directory on the virtualization host was a "stale NFS file handle". I rebooted one of the virtualization hosts (virt1), and tried to make it the SPM. Again, it wouldn't work. Finally, I ended up turning everything into maintenance mode, then activating just it, and I was able to make it the SPM. I was then able to bring everything up. I would have expected ovirt to handle the problem a little more gracefully, and give me more information because I was sweating thinking I had to restore all the VMs!

Stale NFS is on our todo list to handle. Quite challenging.

Thanks..

...

I didn't think when I chose XFS as the filesystem for my virtualization NFS server that I would have to defragment the filesystem manually. This is like the old days of running Norton SpeedDisk to defrag my 386...

We are still not convinced it's an issue - but we'll look into it (and perhaps ask for more stats and data).

Thanks!

...

Y.

Thanks for any help you can provide...

Jason.

...

All 4 virtualization hosts of course had problems since there was no longer any storage.

In the end, it seems like the problem is related to XFS fragmentation...

I read this great blog here:

https://blog.codecentric.de/en/2017/04/xfs-possible-memory-allocation-deadlo... <https://blog.codecentric.de/en/2017/04/xfs-possible-memory-allocation-deadlock-kmem_alloc/>

In short, I tried this:

# xfs_db -r -c "frag -f" /dev/sdb1 actual 4314253, ideal 43107, fragmentation factor 99.00%

Apparently the fragmentation factor doesn't mean much, but the fact that "actual" number of extents is considerably higher than "ideal" extents seems that it may be the problem.

I saw that many of my virtual disks that are written to a lot have, of course, a lot of extents...

For example, on our main web server disk image, there were 247,597 extents alone! I took the web server down, and ran the XFS defrag command on the disk...

# xfs_fsr -v 9a634692-1302-471f-a92e-c978b2b67fd0 9a634692-1302-471f-a92e-c978b2b67fd0 extents before:247597 after:429 DONE 9a634692-1302-471f-a92e-c978b2b67fd0

247,597 before and 429 after! WOW!

Are virtual disks a problem with XFS? Why isn't this memory allocation deadlock issue more prevalent. I do see this article mentioned on many web posts. I don't specifically see any recommendation to *not* use XFS for the data domain though.

I was running CentOS 7.3 on the file server, but before rebooting the server, I upgraded to the latest kernel and CentOS 7.4 in the hopes that if there was a kernel issue, that this would solve it.

I took a few virtual systems down, and ran the defrag on the disks. However, with over 30 virtual systems, I don't really want to do this individually. I was wondering if I could run xfs_fsr on all the disks LIVE? It says in the manual that you can run it live, but I can't see how this would be good when a system is using that disk, and I don't want to deal with major corruption across the board. Any thoughts?

Thanks,

Jason.

_______________________________________________ Users mailing list Users@ovirt.org <mailto:Users@ovirt.org> http://lists.ovirt.org/mailman/listinfo/users <http://lists.ovirt.org/mailman/listinfo/users>

_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users

--------------64D829377F18A02EC6553431 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 8bit <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body text="#000000" bgcolor="#FFFFFF"> <br> <div class="moz-cite-prefix">On 10/02/2017 11:05 AM, Jason Keltz wrote:<br> </div> <blockquote type="cite" cite="mid:b3e2ab23-7bab-ddc2-e230-a650f87a1773@cse.yorku.ca"> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> <div class="moz-cite-prefix">On 10/02/2017 11:00 AM, Yaniv Kaul wrote:<br> </div> <blockquote type="cite" cite="mid:CAJgorsb2ctuEaTpNkzvixsDSjF-_ABH6JDMgw5X03WUgZgbo2A@mail.gmail.com"> <div dir="ltr"><br> <div class="gmail_extra"><br> <div class="gmail_quote">On Mon, Oct 2, 2017 at 5:57 PM, Jason Keltz <span dir="ltr"><<a href="mailto:jas@cse.yorku.ca" target="_blank" moz-do-not-send="true">jas@cse.yorku.ca</a>></span> wrote:<br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"><span class=""> <br> <div class="m_3456688468548054330moz-cite-prefix">On 10/02/2017 10:51 AM, Yaniv Kaul wrote:<br> </div> <blockquote type="cite"> <div dir="ltr"><br> <div class="gmail_extra"><br> <div class="gmail_quote">On Mon, Oct 2, 2017 at 5:14 PM, Jason Keltz <span dir="ltr"><<a href="mailto:jas@cse.yorku.ca" target="_blank" moz-do-not-send="true">jas@cse.yorku.ca</a>></span> wrote:<br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"><span> <br> <div class="m_3456688468548054330m_-6564063642909371047moz-cite-prefix">On 10/02/2017 01:22 AM, Yaniv Kaul wrote:<br> </div> <blockquote type="cite"> <div dir="ltr"><br> <div class="gmail_extra"><br> <div class="gmail_quote">On Mon, Oct 2, 2017 at 5:11 AM, Jason Keltz <span dir="ltr"><<a href="mailto:jas@cse.yorku.ca" target="_blank" moz-do-not-send="true">jas@cse.yorku.ca</a>></span> wrote:<br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi.<br> <br> For my data domain, I have one NFS server with a large RAID filesystem (9 TB).<br> I'm only using 2 TB of that at the moment. Today, my NFS server  hung with<br> the following error:<br> <br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> xfs: possible memory allocation deadlock in kmem_alloc<br> </blockquote> </blockquote> <div><br> </div> <div>Can you share more of the log so we'll see what happened before and after?</div> <div>Y.</div> </div> </div> </div> </blockquote> </span><span class=""> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"> <br> Here is engine-log from yesterday.. the problem started around 14:29 PM.<br> <a class="m_3456688468548054330m_-6564063642909371047moz-txt-link-freetext" href="http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/10012017/engine-log.txt" target="_blank" moz-do-not-send="true">http://www.eecs.yorku.ca/~jas/<wbr>ovirt-debug/10012017/engine-lo<wbr>g.txt</a><br> <br> Here is the vdsm log on one of the virtualization hosts, virt01:<br> <a class="m_3456688468548054330m_-6564063642909371047moz-txt-link-freetext" href="http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/10012017/vdsm.log.2" target="_blank" moz-do-not-send="true">http://www.eecs.yorku.ca/~jas/<wbr>ovirt-debug/10012017/vdsm.log.<wbr>2</a><br> <br> Doing further investigation, I found that the XFS error messages didn't start yesterday.  You'll see they started at the very end of the day on September 23.  See:<br> <br> <a class="m_3456688468548054330m_-6564063642909371047moz-txt-link-freetext" href="http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/messages-20170924" target="_blank" moz-do-not-send="true">http://www.eecs.yorku.ca/~jas/<wbr>ovirt-debug/messages-20170924</a> <br> </div> </blockquote> <div><br> </div> <div>Our storage guys do NOT think it's an XFS fragmentation issue, but we'll be looking at it.</div> <div> </div> </div> </div> </div> </blockquote> </span></div> </blockquote> </div> </div> </div> </blockquote> </span></div> </blockquote> </div> </div> </div> </blockquote> </blockquote> This is an interesting thread to read because the problem sounds quite similar:<br> <br> <a class="moz-txt-link-freetext" href="http://oss.sgi.com/archives/xfs/2016-03/msg00447.html">http://oss.sgi.com/archives/xfs/2016-03/msg00447.html</a><br> <br> In particular, quoted from that:<br> <blockquote type="cite"> <pre>XFS maintains the full extent list for an active inode in memory,</pre> </blockquote> <blockquote type="cite"> <pre>As it is, yes, the memory allocation problem is with the in-core extent tree, and we've known about it for some time. The issue is that as memory gets fragmented, the top level indirection array grows too large to be allocated as a contiguous chunk. When this happens really depends on memory load, uptime and the way the extent tree is being modified. </pre> </blockquote> <br> So in my case, I have a bunch of big XFS disk images for virtual disks.  As the files are big with many extents, keeping all that information in memory at the same time may be the culprit.   Having many extents per se isn't the problem, but having enough memory to be able to store all the information simultaneously may be.  Possible solutions would be to increase the default extent size of the volume (which I'm not sure how to do), defragment the disk, and hence less extents, or potentially add more memory to the file server.  It has 64G.  <br> <br> <blockquote type="cite" cite="mid:b3e2ab23-7bab-ddc2-e230-a650f87a1773@cse.yorku.ca"> <blockquote type="cite" cite="mid:CAJgorsb2ctuEaTpNkzvixsDSjF-_ABH6JDMgw5X03WUgZgbo2A@mail.gmail.com"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"><span class=""> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"><span class=""> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> </div> </div> </div> </blockquote> </span> Hmmm... almost sorry to hear that because that would be easy to "fix"...  <br> <span class=""> <br> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"> <br> They continued on the 24th, then on the 26th... I think there were a few "hangs" on those times that people were complaining about, but we didn't catch the problem.  However, the errors hit big time yesterday at 14:27 PM... see here:<br> <br> <a class="m_3456688468548054330m_-6564063642909371047moz-txt-link-freetext" href="http://www.eecs.yorku.ca/%7Ejas/ovirt-debug/messages-20171001" target="_blank" moz-do-not-send="true">http://www.eecs.yorku.ca/~jas/<wbr>ovirt-debug/messages-20171001</a><br> <br> If you want any other logs, I'm happy to provide them.  I just don't know exactly what to provide.<br> <br> Do you know if I can run the XFS defrag command live? Rather than on a disk by disk, I'd rather just do it on the whole filesystem.  There really aren't that many files since it's just ovirt disk images.  However, I don't understand the implications to running VMs.  I wouldn't want to do anything to create more downtime.<br> </div> </blockquote> <div><br> </div> <div>Should be enough to copy the disks to make them less fragmented.</div> <div> </div> </div> </div> </div> </blockquote> </span> Yes, but this requires downtime.. but there's plenty of additional storage, so this would fix things well.</div> </blockquote> </div> </div> </div> </blockquote> </span></div> </blockquote> <div><br> </div> <div>Live storage migration could be used.</div> <div>Y.</div> <div> </div> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"><span class=""><br> <br> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"> <br> I had upgraded the engine server + 4 virtualization hosts from 4.1.1 to current on September 20 along with upgrading them from CentOS 7.3 to CentOS 7.4.  virtfs, the NFS file server, was running CentOS 7.3 and kernel  vmlinuz-3.10.0-514.16.1.el7.x8<wbr>6_64.  Only yesterday, did I upgrade it to CentOS 7.4 and hence kernel vmlinuz-3.10.0-693.2.2.el7.x86<wbr>_64.<br> <br> I believe the problem is fully XFS related, and not ovirt at all.   Although, I must admit, ovirt didn't help either.  When I rebooted the file server, the iso and export domains were immediately active, but the data domain took quite a long time.  I kept trying to activate it, and it couldn't do it.  I couldn't make a host an SPM.  I found that the data domain directory on the virtualization host was a "stale NFS file handle".  I rebooted one of the virtualization hosts (virt1), and tried to make it the SPM.  Again, it wouldn't work.  Finally, I ended up turning everything into maintenance mode, then activating just it, and I was able to make it the SPM.  I was then able to bring everything up.  I would have expected ovirt to handle the problem a little more gracefully, and give me more information because I was sweating thinking I had to restore all the VMs!<br> </div> </blockquote> <div><br> </div> <div>Stale NFS is on our todo list to handle. Quite challenging.</div> <div> </div> </div> </div> </div> </blockquote> </span> Thanks..<span class=""><br> <br> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"> <br> I didn't think when I chose XFS as the filesystem for my virtualization NFS server that I would have to defragment the filesystem manually.  This is like the old days of running Norton SpeedDisk to defrag my 386...<br> </div> </blockquote> <div><br> </div> <div>We are still not convinced it's an issue - but we'll look into it (and perhaps ask for more stats and data).</div> </div> </div> </div> </blockquote> </span> Thanks! <div> <div class="h5"><br> <br> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <div>Y.</div> <div> </div> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div text="#000000" bgcolor="#FFFFFF"> <br> Thanks for any help you can provide...<span class="m_3456688468548054330HOEnZb"><font color="#888888"><br> <br> Jason.</font></span> <div> <div class="m_3456688468548054330h5"><br> <br> <blockquote type="cite"> <div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote"> <div> </div> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> </blockquote> <br> All 4 virtualization hosts of course had problems since there was no<br> longer any storage.<br> <br> In the end, it seems like the problem is related to XFS fragmentation...<br> <br> I read this great blog here:<br> <br> <a href="https://blog.codecentric.de/en/2017/04/xfs-possible-memory-allocation-deadlo..." rel="noreferrer" target="_blank" moz-do-not-send="true">https://blog.codecentric.de/en<wbr>/2017/04/xfs-possible-memory-a<wbr>llocation-deadlock-kmem_alloc/</a><br> <br> In short, I tried this:<br> <br> # xfs_db -r -c "frag -f" /dev/sdb1<br> actual 4314253, ideal 43107, fragmentation factor 99.00%<br> <br> Apparently the fragmentation factor doesn't mean much, but the fact that<br> "actual" number of extents is considerably higher than "ideal" extents seems that it<br> may be the problem.<br> <br> I saw that many of my virtual disks that are written to a lot have, of course,<br> a lot of extents...<br> <br> For example, on our main web server disk image, there were 247,597<br> extents alone!  I took the web server down, and ran the XFS defrag<br> command on the disk...<br> <br> # xfs_fsr -v 9a634692-1302-471f-a92e-c978b2<wbr>b67fd0<br> 9a634692-1302-471f-a92e-c978b2<wbr>b67fd0<br> extents before:247597 after:429 DONE 9a634692-1302-471f-a92e-c978b2<wbr>b67fd0<br> <br> 247,597 before and 429 after!  WOW!<br> <br> Are virtual disks a problem with XFS?  Why isn't this memory allocation<br> deadlock issue more prevalent.  I do see this article mentioned on many<br> web posts.  I don't specifically see any recommendation to *not* use<br> XFS for the data domain though.<br> <br> I was running CentOS 7.3 on the file server, but before rebooting the server,<br> I upgraded to the latest kernel and CentOS 7.4 in the hopes that if there<br> was a kernel issue, that this would solve it.<br> <br> I took a few virtual systems down, and ran the defrag on the disks.  However,<br> with over 30 virtual systems, I don't really want to do this individually.<br> I was wondering if I could run xfs_fsr on all the disks LIVE?  It says in the<br> manual that you can run it live, but I can't see how this would be good when<br> a system is using that disk, and I don't want to deal with major<br> corruption across the board. Any thoughts?<br> <br> Thanks,<br> <br> Jason.<br>  <br> ______________________________<wbr>_________________<br> Users mailing list<br> <a href="mailto:Users@ovirt.org" target="_blank" moz-do-not-send="true">Users@ovirt.org</a><br> <a href="http://lists.ovirt.org/mailman/listinfo/users" rel="noreferrer" target="_blank" moz-do-not-send="true">http://lists.ovirt.org/mailman<wbr>/listinfo/users</a><br> </blockquote> </div> <br> </div> </div> </blockquote> <br> </div> </div> </div> </blockquote> </div> <br> </div> </div> </blockquote> <br> </div> </div> </div> </blockquote> </div> <br> </div> </div> </blockquote> <br> <br> <fieldset class="mimeAttachmentHeader"></fieldset> <br> <pre wrap="">_______________________________________________ Users mailing list <a class="moz-txt-link-abbreviated" href="mailto:Users@ovirt.org">Users@ovirt.org</a> <a class="moz-txt-link-freetext" href="http://lists.ovirt.org/mailman/listinfo/users">http://lists.ovirt.org/mailman/listinfo/users</a> </pre> </blockquote> <br> </body> </html> --------------64D829377F18A02EC6553431--