[Bugs] [Bug 1416321] read error: Input/ output error but not in brain split stat while file on bricks has different sizes

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1416321

Ravishankar N <ravishankar at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Flags|                            |needinfo?(likunbyl at qq.com)



--- Comment #14 from Ravishankar N <ravishankar at redhat.com> ---
What version of gluster are you using for the clients?

[2017-01-01 16:00:02.990799] E
[afr-self-heal-common.c:233:afr_sh_print_split_brain_log] 0-gvol0-replicate-17:
Unable to self-heal contents of '/gps/up_data/2017-01/02/4.txt' (possible
split-brain). Please delete the file from all but the preferred subvolume.-
Pending matrix:  [ [ 0 0 0 ] [ 0 0 0 ] [ 0 0 0 ] ]
[2017-01-01 16:00:02.991044] E
[afr-self-heal-common.c:2868:afr_log_self_heal_completion_status]
0-gvol0-replicate-17:  backgroung data self heal  failed,   on
/gps/up_data/2017-01/02/4.txt
[2017-01-01 16:00:02.991239] W [afr-open.c:209:afr_open] 0-gvol0-replicate-17:
failed to open as split brain seen, returning EIO
[2017-01-01 16:00:02.991265] W [fuse-bridge.c:2209:fuse_writev_cbk]
0-glusterfs-fuse: 340493338: WRITE => -1 (Input/output error)

>From the client logs, it seems AFR seems to have returned the EIO despite a
pending matrix of all zeroes. But what is noteworthy is that the messages
correspond to an older version of afr (i.e. afr-v1 in glusterfs-3.5 or older).

1. What version of gluster are you using for the clients?
2. It is recommended to update all the clients also to the same version (3.8.x)
as the servers.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.