[Gluster-users] split-brain sanlock ids automation

[Ravishankar N]

On 05/26/2015 08:48 AM, maokx at sina.com wrote:
> Hi all:
> I want to solve the problem of ids spit-brain automation.
> This problem is caused by network interruption.
> My hand is such a solution:
> find /data/data1/ -samefile 
> /data/data1/3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids -print -delete
> The problem is caused by sanlock,for example:sanlock add_lockspace -s 
> lockspace1:1:/dom_md/ids :0
> Now I want to automation to solve this problem


You can use the the gluster CLI commands (glusterfs 3.6 onwards) (or) 
the get/setfattr commands from the mount (glusterfs 3.7 onwards) to heal 
files that are in split-brain. Usage can be found at 
https://github.com/gluster/glusterfs/blob/master/doc/features/heal-info-and-split-brain-resolution.md
If you are using ovirt (and hence the sanlock file) with gluster, you 
can achieve better split-brain protection using replica 3 volumes with 
cluster.quorum-type set to auto.

Hope that helps,
Ravi

> *my log is:*
>
> [root at www ~]# tail -f /var/log/glusterfs/rhev-data-center-mnt-glusterSD-192.168.7.246\:_pool01.log 
>
>
> [2015-05-25 06:37:33.613126] W [page.c:991:__ioc_page_error] 0-pool01-io-cache: page error for page = 0x7f4b9801d2d0 & waitq = 0x7f4b9801dd70
>
> [2015-05-25 06:37:33.613157] W [fuse-bridge.c:2089:fuse_readv_cbk] 0-glusterfs-fuse: 246963: READ => -1 (Input/output error)
>
> [2015-05-25 06:37:37.422720] E [afr-self-heal-common.c:197:afr_sh_print_split_brain_log] 0-pool01-replicate-0: Unable to self-heal contents of '/3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids' (possible split-brain). Please delete the file from all but the preferred subvolume.- Pending matrix:  [ [ 0 4 ] [ 4 0 ] ]
>
> [2015-05-25 06:37:37.422981] E [afr-self-heal-common.c:2212:afr_self_heal_completion_cbk] 0-pool01-replicate-0: background  data self-heal failed on /3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids
>
> [2015-05-25 06:37:37.423231] W [page.c:991:__ioc_page_error] 0-pool01-io-cache: page error for page = 0x7f4b9800c520 & waitq = 0x7f4b9801aff0
>
> [2015-05-25 06:37:37.423259] W [fuse-bridge.c:2089:fuse_readv_cbk] 0-glusterfs-fuse: 246978: READ => -1 (Input/output error)
>
> [2015-05-25 06:37:43.650389] E [afr-self-heal-common.c:197:afr_sh_print_split_brain_log] 0-pool01-replicate-0: Unable to self-heal contents of '/3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids' (possible split-brain). Please delete the file from all but the preferred subvolume.- Pending matrix:  [ [ 0 4 ] [ 4 0 ] ]
>
> [2015-05-25 06:37:43.650740] E [afr-self-heal-common.c:2212:afr_self_heal_completion_cbk] 0-pool01-replicate-0: background  data self-heal failed on 
> /3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids
>
> [2015-05-25 06:37:43.650994] W [page.c:991:__ioc_page_error] 0-pool01-io-cache: page error for page = 0x7f4b9802adc0 & waitq = 0x7f4b98017e50
>
> [2015-05-25 06:37:43.651021] W [fuse-bridge.c:2089:fuse_readv_cbk] 0-glusterfs-fuse: 246997: READ => -1 (Input/output error)
>
> [2015-05-25 06:37:50.931622] E [afr-self-heal-common.c:197:afr_sh_print_split_brain_log] 0-pool01-replicate-0: Unable to self-heal contents of '/3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids' (possible split-brain). Please delete the file from all but the preferred subvolume.- Pending matrix:  [ [ 0 4 ] [ 4 0 ] ]
>
> [2015-05-25 06:37:50.931906] E [afr-self-heal-common.c:2212:afr_self_heal_completion_cbk] 0-pool01-replicate-0: background  data self-heal failed on /3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids
>
> [2015-05-25 06:37:50.932211] W [page.c:991:__ioc_page_error] 0-pool01-io-cache: page error for page = 0x7f4b980211e0 & waitq = 0x7f4b980065e0
>
> [2015-05-25 06:37:50.932240] W [fuse-bridge.c:2089:fuse_readv_cbk] 0-glusterfs-fuse: 247012: READ => -1 (Input/output error)
>
> [2015-05-25 06:37:53.688445] E [afr-self-heal-common.c:197:afr_sh_print_split_brain_log] 0-pool01-replicate-0: Unable to self-heal contents of '/3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids' (possible split-brain). Please delete the file from all but the preferred subvolume.- Pending matrix:  [ [ 0 4 ] [ 4 0 ] ]
>
> [2015-05-25 06:37:53.688821] E [afr-self-heal-common.c:2212:afr_self_heal_completion_cbk] 0-pool01-replicate-0: background  data self-heal failed on /3b9e1af1-f452-4ed4-9cbf-a38d8f7a395e/dom_md/ids
>
> [2015-05-25 06:37:53.689128] W [page.c:991:__ioc_page_error] 0-pool01-io-cache: page error for page = 0x7f4b9800acb0 & waitq = 0x7f4b9802b4d0
>
> [2015-05-25 06:37:53.689152] W [fuse-bridge.c:2089:fuse_readv_cbk] 0-glusterfs-fuse: 247031: READ => -1 (Input/output error)
>
> *I am such a environment:*
>
>
> ------------------------------------------------------------------------
> maokx at sina.com
>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://www.gluster.org/mailman/listinfo/gluster-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20150526/01e6c7a0/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 82928 bytes
Desc: not available
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20150526/01e6c7a0/attachment.jpe>