Hi mabi, there are a couple of AFR patches from master that I'm
currently back porting to the 3.12 branch:
afr: heal gfids when file is not present on all bricks
afr: don't update readables if inode refresh failed on all children
afr: fix bug-1363721.t failure
afr: add quorum checks in pre-op
afr: don't treat all cases all bricks being blamed as split-brain
afr: capture the correct errno in post-op quorum check
afr: add quorum checks in post-op
Many of these help make the transaction code more robust by fixing
various corner cases. It would be great if you can wait for the next
3.12 minor release (3.12.12 ?) and upgrade to that build and see if the
issues go away.
Note: CC'ing Karthik and Jiffin for their help in reviewing and merging
the backports for the above patches.
Thanks,
Ravi
On 07/04/2018 06:51 PM, mabi wrote:
Hello,
I just wanted to let you know that last week I have upgraded my two replica nodes from Debian 8 to Debian 9 so now all my 3 nodes (including aribter) are running Debian 9 with a Linux 4 kernel.
Unfortunately I still have the exact same issue. Another detail I might have not mentioned yet is that I have quotas enabled on this volume, I don't really know if that is relevant but who knows...
As a reminder here is what happens on the client side which has the volume mounted via FUSE (take earlier today from the /var/log/glusterfs/mnt-myvol-private.log logfile). Note here that in this specific case it's only one single file who had this issue.
[2018-07-04 08:23:49.314252] E [MSGID: 109089] [dht-helper.c:1481:dht_migration_complete_check_task] 0-myvol-private-dht: failed to open the fd (0x7fccb00a5120, flags=0100000) on file /dir1/data/dir2/files_encryption/keys/files/dir3/dir4/dir5/dir6/dir7/OC_DEFAULT_MODULE/file.shareKey @ myvol-replicate-0 [Input/output error]
[2018-07-04 08:23:49.328712] W [MSGID: 108027] [afr-common.c:2821:afr_discover_done] 0-myvol-private-replicate-0: no read subvols for /dir1/data/dir2/files_encryption/keys/files/dir3/dir4/dir5/dir6/dir7/OC_DEFAULT_MODULE/file.shareKey
[2018-07-04 08:23:49.330749] W [fuse-bridge.c:779:fuse_truncate_cbk] 0-glusterfs-fuse: 55916791: TRUNCATE() /dir1/data/dir2/files_encryption/keys/files/dir3/dir4/dir5/dir6/dir7/OC_DEFAULT_MODULE/file.shareKey => -1 (Input/output error)
Best regards,
M.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On June 22, 2018 4:44 PM, mabi <mabi@xxxxxxxxxxxxx> wrote:
Hi,
Now that this issue has happened a few times I noticed a few things which might be helpful for debugging:
- This problem happens when files are uploaded via a cloud app called Nextcloud where the files are encrypted by the app itself on the server side (PHP code) but only rarely and randomly.
- It does not seem to happen with Nextcloud installation which does not have server side encryption enabled.
- When this happens both first and second node of the replica have 120k of context switches and 25k interrupts, the arbiter node 30k context switches/20k interrupts. No nodes are overloaded, there is no io/wait and no network issues or disconnections.
- All of the problematic files to heal have spaces in one of their sub-directories (might be totally irrelevant).
If that's of any use my two replica nodes are Debian 8 physical servers with ZFS as file system for the bricks and the arbiter is a Debian 9 virtual machine with XFS as file system for the brick. To mount the volume I use a glusterfs fuse mount on the web server which has Nextcloud running.
Regards,
M.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On May 25, 2018 5:55 PM, mabi mabi@xxxxxxxxxxxxx wrote:
Thanks Ravi. Let me know when you have time to have a look. It sort of happens around once or twice per week but today it was 24 files in one go which are unsynched and where I need to manually reset the xattrs on the arbiter node.
By the way on this volume I use quotas which I set on specifc directories, I don't know if this is relevant or not but thought I would just mention.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On May 23, 2018 9:25 AM, Ravishankar N ravishankar@xxxxxxxxxx wrote:
On 05/23/2018 12:47 PM, mabi wrote:
Hello,
I just wanted to ask if you had time to look into this bug I am encountering and if there is anything else I can do?
For now in order to get rid of these 3 unsynched files shall I do the same method that was suggested to me in this thread?
Sorry Mabi, I haven't had a chance to dig deeper into this. The
workaround of resetting xattrs should be fine though.
Thanks,
Ravi
Thanks,
Mabi
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On May 17, 2018 11:07 PM, mabi mabi@xxxxxxxxxxxxx wrote:
Hi Ravi,
Please fine below the answers to your questions
1. I have never touched the cluster.quorum-type option. Currently it is set as following for this volume:
Option Value
cluster.quorum-type none
2. The .shareKey files are not supposed to be empty. They should be 512 bytes big and contain binary data (PGP Secret Sub-key). I am not in a position to say why it is in this specific case only 0 bytes and if it is the fault of the software (Nextcloud) or GlusterFS. I can just say here that I have another file server which is a simple NFS server with another Nextcloud installation and there I never saw any 0 bytes .shareKey files being created.
3. It seems to be quite random and I am not the person who uses the Nextcloud software so I can't say what it was doing at that specific time but I guess uploading files or moving files around. Basically I use GlusterFS to store the files/data of the Nextcloud web application where I have it mounted using a fuse mount (mount -t glusterfs).
Regarding the logs I have attached the mount log file from the client and below are the relevant log entries from the brick log file of all 3 nodes. Let me know if you need any other log files. Also if you know any "log file sanitizer tool" which can replace sensitive file names with random file names in log files that would like to use it as right now I have to do that manually.
NODE 1 brick log:
[2018-05-15 06:54:20.176679] E [MSGID: 113015] [posix.c:1211:posix_opendir] 0-myvol-private-posix: opendir failed on /data/myvol-private/brick/cloud/data/admin/files_encryption/keys/files/dir/dir/anotherdir/dir/OC_DEFAULT_MODULE [No such file or directory]
NODE 2 brick log:
[2018-05-15 06:54:20.176415] E [MSGID: 113015] [posix.c:1211:posix_opendir] 0-myvol-private-posix: opendir failed on /data/myvol-private/brick/cloud/data/admin/files_encryption/keys/files/dir/dir/anotherdir/dir/OC_DEFAULT_MODULE [No such file or directory]
NODE 3 (arbiter) brick log:
[2018-05-15 06:54:19.898981] W [MSGID: 113103] [posix.c:285:posix_lookup] 0-myvol-private-posix: Found stale gfid handle /srv/glusterfs/myvol-private/brick/.glusterfs/f0/65/f065a5e7-ac06-445f-add0-83acf8ce4155, removing it. [Stale file handle]
[2018-05-15 06:54:20.056196] W [MSGID: 113103] [posix.c:285:posix_lookup] 0-myvol-private-posix: Found stale gfid handle /srv/glusterfs/myvol-private/brick/.glusterfs/8f/a1/8fa15dbd-cd5c-4900-b889-0fe7fce46a13, removing it. [Stale file handle]
[2018-05-15 06:54:20.172823] I [MSGID: 115056] [server-rpc-fops.c:485:server_rmdir_cbk] 0-myvol-private-server: 14740125: RMDIR /cloud/data/admin/files_encryption/keys/files/dir/dir/anotherdir/dir/OC_DEFAULT_MODULE (f065a5e7-ac06-445f-add0-83acf8ce4155/OC_DEFAULT_MODULE), client: nextcloud.domain.com-7972-2018/05/10-20:31:46:163206-myvol-private-client-2-0-0, error-xlator: myvol-private-posix [Directory not empty]
[2018-05-15 06:54:20.190911] I [MSGID: 115056] [server-rpc-fops.c:485:server_rmdir_cbk] 0-myvol-private-server: 14740141: RMDIR /cloud/data/admin/files_encryption/keys/files/dir/dir/anotherdir/dir (72a1613e-2ac0-48bd-8ace-f2f723f3796c/2016.03.15 AVB_Photovoltaik-Versicherung 2013.pdf), client: nextcloud.domain.com-7972-2018/05/10-20:31:46:163206-myvol-private-client-2-0-0, error-xlator: myvol-private-posix [Directory not empty]
Best regards,
Mabi
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On May 17, 2018 7:00 AM, Ravishankar N ravishankar@xxxxxxxxxx wrote:
Hi mabi,
Some questions:
-Did you by any chance change the cluster.quorum-type option from the
default values?
-Is filename.shareKey supposed to be any empty file? Looks like the file
was fallocated with the keep-size option but never written to. (On the 2
data bricks, stat output shows Size =0, but non zero Blocks and yet a
'regular empty file').
-Do you have some sort of a reproducer/ steps that you perform when the
issue occurs? Please also share the logs from all 3 nodes and the client(s).
Thanks,
Ravi
On 05/15/2018 05:26 PM, mabi wrote:
Thank you Ravi for your fast answer. As requested you will find below the "stat" and "getfattr" of one of the files and its parent directory from all three nodes of my cluster.
NODE 1:
File: ‘/data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey’
Size: 0 Blocks: 38 IO Block: 131072 regular empty file
Device: 23h/35d Inode: 744413 Links: 2
Access: (0644/-rw-r--r--) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-05-15 08:54:20.296048887 +0200
Modify: 2018-05-15 08:54:20.296048887 +0200
Change: 2018-05-15 08:54:20.340048505 +0200
Birth: -
File: ‘/data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/’
Size: 8 Blocks: 74 IO Block: 131072 directory
Device: 23h/35d Inode: 744410 Links: 2
Access: (0755/drwxr-xr-x) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-04-25 09:41:24.276780766 +0200
Modify: 2018-05-15 08:54:20.392048056 +0200
Change: 2018-05-15 08:54:20.392048056 +0200
Birth: -
file: data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey
========================================================================================================================
trusted.gfid=0x3b6c722cd6c64a4180fa028809671d63
trusted.gfid2path.9cb852a48fe5e361=0x38666131356462642d636435632d343930302d623838392d3066653766636534366131332f6e6361646d696e6973747261746f722e73686172654b6579
trusted.glusterfs.quota.8fa15dbd-cd5c-4900-b889-0fe7fce46a13.contri.1=0x00000000000000000000000000000001
trusted.pgfid.8fa15dbd-cd5c-4900-b889-0fe7fce46a13=0x00000001
file: data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/
=======================================================================================================
trusted.gfid=0x8fa15dbdcd5c4900b8890fe7fce46a13
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.f065a5e7-ac06-445f-add0-83acf8ce4155.contri.1=0x000000000000060000000000000000060000000000000001
trusted.glusterfs.quota.size.1=0x000000000000060000000000000000060000000000000001
NODE 2:
File: ‘/data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey’
Size: 0 Blocks: 38 IO Block: 131072 regular empty file
Device: 24h/36d Inode: 5428150 Links: 2
Access: (0644/-rw-r--r--) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-05-15 08:54:20.294280254 +0200
Modify: 2018-05-15 08:54:20.294280254 +0200
Change: 2018-05-15 08:54:20.338279576 +0200
Birth: -
File: ‘/data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/’
Size: 8 Blocks: 74 IO Block: 131072 directory
Device: 24h/36d Inode: 5428147 Links: 2
Access: (0755/drwxr-xr-x) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-04-25 09:41:24.276780766 +0200
Modify: 2018-05-15 08:54:20.394278717 +0200
Change: 2018-05-15 08:54:20.394278717 +0200
Birth: -
file: data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey
========================================================================================================================
trusted.gfid=0x3b6c722cd6c64a4180fa028809671d63
trusted.gfid2path.9cb852a48fe5e361=0x38666131356462642d636435632d343930302d623838392d3066653766636534366131332f6e6361646d696e6973747261746f722e73686172654b6579
trusted.glusterfs.quota.8fa15dbd-cd5c-4900-b889-0fe7fce46a13.contri.1=0x00000000000000000000000000000001
trusted.pgfid.8fa15dbd-cd5c-4900-b889-0fe7fce46a13=0x00000001
file: data/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/
=======================================================================================================
trusted.gfid=0x8fa15dbdcd5c4900b8890fe7fce46a13
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.f065a5e7-ac06-445f-add0-83acf8ce4155.contri.1=0x000000000000060000000000000000060000000000000001
trusted.glusterfs.quota.size.1=0x000000000000060000000000000000060000000000000001
NODE 3 (arbiter):
File: /srv/glusterfs/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey
Size: 0 Blocks: 8 IO Block: 4096 regular empty file
Device: ca11h/51729d Inode: 271434295 Links: 2
Access: (0644/-rw-r--r--) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-04-25 09:41:24.322527555 +0200
Modify: 2018-04-25 09:41:24.322527555 +0200
Change: 2018-05-15 08:54:20.343667380 +0200
Birth: -
File: /srv/glusterfs/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/
Size: 8192 Blocks: 24 IO Block: 4096 directory
Device: ca11h/51729d Inode: 271434288 Links: 2
Access: (0755/drwxr-xr-x) Uid: (20936/ UNKNOWN) Gid: (20936/ UNKNOWN)
Access: 2018-04-25 09:41:24.276780766 +0200
Modify: 2018-05-15 08:54:20.391667997 +0200
Change: 2018-05-15 08:54:20.395668048 +0200
Birth: -
file: srv/glusterfs/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/filename.shareKey
=================================================================================================================================
trusted.afr.dirty=0x000000000000000000000000
trusted.afr.myvolume-private-client-0=0x000000010000000000000000
trusted.afr.myvolume-private-client-1=0x000000010000000000000000
trusted.gfid=0x3b6c722cd6c64a4180fa028809671d63
trusted.gfid2path.9cb852a48fe5e361=0x38666131356462642d636435632d343930302d623838392d3066653766636534366131332f6e6361646d696e6973747261746f722e73686172654b6579
trusted.glusterfs.quota.8fa15dbd-cd5c-4900-b889-0fe7fce46a13.contri.1=0x00000000000000000000000000000001
trusted.pgfid.8fa15dbd-cd5c-4900-b889-0fe7fce46a13=0x00000001
file: srv/glusterfs/myvolume-private/brick/dir1/dir2/dir3/dir4/dir5/dir6/dir7/dir8/dir9/dir10/OC_DEFAULT_MODULE/
================================================================================================================
trusted.afr.myvolume-private-client-0=0x000000000000000000000000
trusted.afr.myvolume-private-client-1=0x000000000000000000000000
trusted.gfid=0x8fa15dbdcd5c4900b8890fe7fce46a13
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.f065a5e7-ac06-445f-add0-83acf8ce4155.contri.1=0x000000000000000000000000000000060000000000000001
trusted.glusterfs.quota.size.1=0x000000000000000000000000000000060000000000000001
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On May 15, 2018 10:52 AM, Ravishankar N ravishankar@xxxxxxxxxx wrote:
On 05/15/2018 12:38 PM, mabi wrote:
Dear all,
I have upgraded my replica 3 GlusterFS cluster (and clients) last Friday from 3.12.7 to 3.12.9 in order to fix this bug but unfortunately I notice that I still have exactly the same problem as initially posted in this thread.
It looks like this bug is not resolved as I just got right now 3 unsynched files on my arbiter node like I used to do before upgrading. This problem started since I upgraded to 3.12.7...
Could you provide the stat and 'getfattr -d -m . - hex
brick/path/to/file' outputs of one of these files and also the
corresponding parent directory from all 3 bricks?
Thanks,
Ravi
Thank you very much in advance for your advise.
Best regards,
Mabi
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On April 9, 2018 2:31 PM, Ravishankar N ravishankar@xxxxxxxxxx wrote:
On 04/09/2018 05:54 PM, Dmitry Melekhov wrote:
09.04.2018 16:18, Ravishankar N пишет:
On 04/09/2018 05:40 PM, mabi wrote:
Again thanks that worked and I have now no more unsynched files.
You mentioned that this bug has been fixed in 3.13, would it be
possible to backport it to 3.12? I am asking because 3.13 is not a
long-term release and as such I would not like to have to upgrade to
3.13.
I don't think there will be another 3.12 release.
Why not? It is LTS, right?
My bad. Just checked the schedule [1], and you are right. It is LTM.
[1] https://www.gluster.org/release-schedule/
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-users
