[Gluster-users] mirror-0 volume issues

[Vladislav Tchernev]

Hello List,
I have two gluster 3.0.5 servers exporting a couple of mirror-0 volumes to a 
bunch of clients. One of the volumes stopped replicating while the other 
volumes are still working well. A client can mount the trouble volume from 
both gluster servers but a write is going only to the server that the client 
is mounting the volume from and not to the other server. Looking in the log 
files I noticed the following behavior:

When a client mounts the volume from server A, I am getting the following 
messages in the log files:

1. Log file on server A:
[2010-12-20 11:21:40] N [server-protocol.c:5852:mop_setvolume] server-tcp: 
accepted client from Client_IP:1022
[2010-12-20 11:21:40] N [server-protocol.c:5852:mop_setvolume] server-tcp: 
accepted client from Client_IP:1008

2. Log file on server B:
[2010-12-20 11:59:41] N [server-helpers.c:842:server_connection_destroy] 
server-tcp: destroyed connection of ClentName-5678-2010/12/20-14:59:41:630525-
gluster2-volume100
[2010-12-20 11:59:52] N [server-protocol.c:6788:notify] server-tcp: 
ClientIP:1016 disconnected
[2010-12-20 11:59:52] N [server-helpers.c:842:server_connection_destroy] 
server-tcp: destroyed connection of 
ClientName-5678-2010/12/20-14:59:41:630525-gluster2-volume100
[2010-12-20 11:59:52] N [server-protocol.c:6788:notify] server-tcp: 
ClientIP:1015 disconnected
[2010-12-20 11:59:52] N [server-helpers.c:842:server_connection_destroy] 
server-tcp: destroyed connection of 
ClientName-5678-2010/12/20-14:59:41:630525-gluster2-volume100
[2010-12-20 12:00:02] N [server-protocol.c:6788:notify] server-tcp: 
ClientIP:1014 disconnected
.....................................................................
.....................................................................

And the above messages are flooding the log file on server B

3. Log file on the client:
.................
[2010-12-20 14:11:46] N [glusterfsd.c:1409:main] glusterfs: Successfully 
started
[2010-12-20 14:11:46] N [client-protocol.c:6288:client_setvolume_cbk] 
gluster1-volume100: Connected to ServerA_IP:6996, attached to remote volume 
'brick-volume100'.
[2010-12-20 14:11:46] N [afr.c:2648:notify] mirror-0: Subvolume 'gluster1-
volume100' came back up; going online.
[2010-12-20 14:11:46] N [client-protocol.c:6288:client_setvolume_cbk] 
gluster1-volume100: Connected to ServerA_IP:6996, attached to remote volume 
'brick-volume100'.
[2010-12-20 14:11:46] N [afr.c:2648:notify] mirror-0: Subvolume 'gluster1-
volume100' came back up; going online.
[2010-12-20 14:11:46] N [fuse-bridge.c:2953:fuse_init] glusterfs-fuse: FUSE 
inited with protocol versions: glusterfs 7.13 kernel 7.13
...........................................
...........................................
[2010-12-20 14:13:05] E [afr.c:107:afr_set_split_brain] mirror-0: invalid 
argument: inode
...................................



The opposite behavior will be observed if the client mounts the volume from 
server B insetad of server A.

I know that there is a new version of glusterfs and I am planning to migrate 
it soon but in the mean time I have to fix the current setup.

Thank you in advance!!!

Regards,
Vlad