[Gluster-users] gluster 3.7.11 qemu+libgfapi problem

[Dmitry Melekhov]

Hello!

I have 3 servers setup- centos 7 and gluster 3.7.11
and don't know did it work with previous versions or not...


Volume is replicated 3.

If I shutdown port on switch for one of nodes, then qemu can't start, 
because it can't connects to gluster:



[2016-04-26 10:51:53.881654] I [MSGID: 114057] 
[client-handshake.c:1437:select_server_supported_programs] 
0-pool-client-7: Using Program GlusterFS 3.3, Num (1298437), Version (330)
[2016-04-26 10:51:53.882271] I [MSGID: 114046] 
[client-handshake.c:1213:client_setvolume_cbk] 0-pool-client-7: 
Connected to pool-client-7, attached to remote volume '/wall/pool/brick'.
[2016-04-26 10:51:53.882299] I [MSGID: 114047] 
[client-handshake.c:1224:client_setvolume_cbk] 0-pool-client-7: Server 
and Client lk-version numbers are not same, reopening the fds
[2016-04-26 10:51:53.882620] I [MSGID: 114035] 
[client-handshake.c:193:client_set_lk_version_cbk] 0-pool-client-7: 
Server lk version = 1
[2016-04-26 10:51:55.373983] E [socket.c:2279:socket_connect_finish] 
0-pool-client-8: connection to 192.168.22.28:24007 failed (No route to host)
[2016-04-26 10:51:55.416522] I [MSGID: 108031] 
[afr-common.c:1900:afr_local_discovery_cbk] 0-pool-replicate-0: 
selecting local read_child pool-client-6
[2016-04-26 10:51:55.416919] I [MSGID: 104041] 
[glfs-resolve.c:869:__glfs_active_subvol] 0-pool: switched to graph 
66617468-6572-2d35-3334-372d32303136 (0)
qemu: terminating on signal 15 from pid 9767
[2016-04-26 10:53:36.418693] I [MSGID: 114021] [client.c:2115:notify] 
0-pool-client-6: current graph is no longer active, destroying rpc_client
[2016-04-26 10:53:36.418802] I [MSGID: 114021] [client.c:2115:notify] 
0-pool-client-7: current graph is no longer active, destroying rpc_client
[2016-04-26 10:53:36.418840] I [MSGID: 114021] [client.c:2115:notify] 
0-pool-client-8: current graph is no longer active, destroying rpc_client
[2016-04-26 10:53:36.418870] I [MSGID: 114018] 
[client.c:2030:client_rpc_notify] 0-pool-client-6: disconnected from 
pool-client-6. Client process will keep trying to connect to glusterd 
until brick's port is avai
lable
[2016-04-26 10:53:36.418880] I [MSGID: 114018] 
[client.c:2030:client_rpc_notify] 0-pool-client-7: disconnected from 
pool-client-7. Client process will keep trying to connect to glusterd 
until brick's port is avai
lable
[2016-04-26 10:53:36.418949] W [MSGID: 108001] 
[afr-common.c:4090:afr_notify] 0-pool-replicate-0: Client-quorum is not met
[2016-04-26 10:53:36.419002] E [MSGID: 108006] 
[afr-common.c:4046:afr_notify] 0-pool-replicate-0: All subvolumes are 
down. Going offline until atleast one of them comes back up.


192.168.22.28 is node , which is not available.

I don't see any errors in bricks logs, only
[2016-04-26 10:53:41.807032] I [dict.c:473:dict_get] 
(-->/lib64/libglusterfs.so.0(default_getxattr_cbk+0xac) [0x7f7405415cbc] 
-->/usr/lib64/glusterfs/3.7.11/xlator/features/marker.so(marker_getxattr_cbk+0xa7) 
[0x
7f73f59da917] -->/lib64/libglusterfs.so.0(dict_get+0xac) 
[0x7f74054060fc] ) 0-dict: !this || key=() [Invalid argument]

But I guess it is not related.


Could you tell me what can cause this problem ?

Thank you!