[Gluster-devel] Losing connection to bricks, Gluster processes restarting
Logan Barfield
lbarfield at tqhosting.com
Tue Feb 2 16:55:31 UTC 2016
Not historically, but we are using bonding for replication between the
servers. It's been stable for at least 6 months, but it's possible that
one of the links in the bond is failing or something.
Would this type of restart be triggered by a loss of communication between
bricks in a replica set? It seems like it would defeat one of the points
of having a replicated volume if that were the case.
Thank You,
Logan Barfield
Tranquil Hosting
On Tue, Feb 2, 2016 at 12:02 AM, Atin Mukherjee <amukherj at redhat.com> wrote:
> Initially I was suspecting about server-quorum be the culprit which is
> not the case. By any chance is your network flaky?
>
> On 02/01/2016 10:33 PM, Logan Barfield wrote:
> > Volume Name: data02
> > Type: Replicate
> > Volume ID: 1c8928b1-f49e-4950-be06-0f8ce5adf870
> > Status: Started
> > Number of Bricks: 1 x 2 = 2
> > Transport-type: tcp
> > Bricks:
> > Brick1: gluster-stor01:/export/data/brick02 <-- 10.1.1.10
> > Brick2: gluster-stor02:/export/data/brick02 <-- 10.1.1.11
> > Options Reconfigured:
> > server.event-threads: 5
> > client.event-threads: 11
> > geo-replication.indexing: on
> > geo-replication.ignore-pid-check: on
> > changelog.changelog: on
> > server.statedump-path: /tmp
> > server.outstanding-rpc-limit: 128
> > performance.io-thread-count: 64
> > performance.nfs.read-ahead: on
> > performance.nfs.io-cache: on
> > performance.nfs.quick-read: on
> > performance.cache-max-file-size: 1MB
> > performance.client-io-threads: on
> > cluster.lookup-optimize: on
> > performance.cache-size: 1073741824
> > performance.write-behind-window-size: 4MB
> > performance.nfs.write-behind-window-size: 4MB
> > performance.read-ahead: off
> > performance.nfs.stat-prefetch: on
> >
> >
> > Status of volume: data02
> > Gluster process TCP Port RDMA Port Online
> Pid
> >
> ------------------------------------------------------------------------------
> > Brick gluster-stor01:/export/data/brick02 49153 0 Y
> > 17411
> > Brick gluster-stor02:/export/data/brick02 49155 0 Y
> > 4717
> > NFS Server on localhost 2049 0 Y
> > 17395
> > Self-heal Daemon on localhost N/A N/A Y
> > 17405
> > NFS Server on gluster-stor02 2049 0 Y
> > 4701
> > Self-heal Daemon on gluster-stor02 N/A N/A Y
> > 4712
> >
> > Task Status of Volume data02
> >
> ------------------------------------------------------------------------------
> > There are no active volume tasks
> >
> >
> >
> > Note that this problem was occurring with the same frequency before we
> > added all of the volume options above. We were running defaults up
> > until last week, and changing them had no impact on this particular
> problem.
> >
> >
> >
> >
> > Thank You,
> >
> > Logan Barfield
> > Tranquil Hosting
> >
> > On Fri, Jan 29, 2016 at 9:28 PM, Atin Mukherjee <amukherj at redhat.com
> > <mailto:amukherj at redhat.com>> wrote:
> >
> > Could you paste output of gluster volume info?
> >
> > ~Atin
> >
> > On 01/29/2016 11:59 PM, Logan Barfield wrote:
> > > We're running a fairly large 2-replica volume across two servers.
> The
> > > volume is approximately 20TB of small 1K-4MB files. The volume is
> > > exported via NFS, and mounted remotely by two clients.
> > >
> > > For the past few weeks the Gluster brick processes have been
> randomly
> > > restarting. Luckily they've been doing so at non-peak times, so we
> > > didn't notice until our monitoring checks happened to pick up on
> > zombied
> > > 'glusterfs' process.
> > >
> > > From the logs it looks like something is blocking communication to
> the
> > > brick processes, and Gluster automatically restarts everything to
> > > compensate. I've so far not been able to figure out the
> > underlying cause.
> > >
> > > I've included log snippets from 'glustershd.log' and
> > > 'etc-glusterfs-glusterd.vol.log' here. If anyone can provide some
> > > insight into the issue it would be greatly appreciated. I'll also
> be
> > > happy to provide any further details as needed.
> > >
> > >
> > > [2016-01-29 05:03:47.039886] I [MSGID: 106144]
> > > [glusterd-pmap.c:274:pmap_registry_remove] 0-pmap: removing brick
> > > /export/data/brick02 on port 49155
> > > [2016-01-29 05:03:47.075521] W [socket.c:588:__socket_rwv]
> > 0-management:
> > > readv on /var/run/gluster/53a233b05f5d4be45dc94391bc3ebfe5.socket
> > failed
> > > (No data available)
> > > [2016-01-29 05:03:47.078282] I [MSGID: 106005]
> > > [glusterd-handler.c:4908:__glusterd_brick_rpc_notify] 0-management:
> > > Brick gluster-stor02:/export/data/brick02 has disconnected from
> > glusterd.
> > > [2016-01-29 05:03:47.149161] W [glusterfsd.c:1236:cleanup_and_exit]
> > > (-->/lib64/libpthread.so.0() [0x3e47a079d1]
> > > -->/usr/sbin/glusterd(glusterfs_sigwaiter+0xcd) [0x405e6d]
> > > -->/usr/sbin/glusterd(cleanup_and_exit+0x65) [0x4059d5] ) 0-: recei
> > > ved signum (15), shutting down
> > > [2016-01-29 05:03:54.067012] I [MSGID: 100030]
> > [glusterfsd.c:2318:main]
> > > 0-/usr/sbin/glusterd: Started running /usr/sbin/glusterd version
> 3.7.6
> > > (args: /usr/sbin/glusterd --pid-file=/var/run/glusterd.pid)
> > > [2016-01-29 05:03:54.071901] I [MSGID: 106478]
> [glusterd.c:1350:init]
> > > 0-management: Maximum allowed open file descriptors set to 65536
> > > [2016-01-29 05:03:54.071935] I [MSGID: 106479]
> [glusterd.c:1399:init]
> > > 0-management: Using /var/lib/glusterd as working directory
> > > [2016-01-29 05:03:54.075655] E
> > [rpc-transport.c:292:rpc_transport_load]
> > > 0-rpc-transport: /usr/lib64/glusterfs/3.7.6/rpc-transport/rdma.so:
> > > cannot open shared object file: No such file or directory
> > > [2016-01-29 05:03:54.075672] W
> > [rpc-transport.c:296:rpc_transport_load]
> > > 0-rpc-transport: volume 'rdma.management': transport-type 'rdma'
> > is not
> > > valid or not found on this machine
> > > [2016-01-29 05:03:54.075680] W
> [rpcsvc.c:1597:rpcsvc_transport_create]
> > > 0-rpc-service: cannot create listener, initing the transport failed
> > > [2016-01-29 05:03:54.075687] E [MSGID: 106243]
> [glusterd.c:1623:init]
> > > 0-management: creation of 1 listeners failed, continuing with
> > succeeded
> > > transport
> > > [2016-01-29 05:03:55.869717] I [MSGID: 106513]
> > > [glusterd-store.c:2047:glusterd_restore_op_version] 0-glusterd:
> > > retrieved op-version: 30702
> > > [2016-01-29 05:03:55.995747] I [MSGID: 106498]
> > > [glusterd-handler.c:3579:glusterd_friend_add_from_peerinfo]
> > > 0-management: connect returned 0
> > > [2016-01-29 05:03:55.995866] I
> > [rpc-clnt.c:984:rpc_clnt_connection_init]
> > > 0-management: setting frame-timeout to 600
> > > [2016-01-29 05:03:56.000937] I [MSGID: 106544]
> > > [glusterd.c:159:glusterd_uuid_init] 0-management: retrieved UUID:
> > > 9b103ea8-d248-44fc-8f80-3e87f7c4971c
> > > Final graph:
> > >
> >
> +------------------------------------------------------------------------------+
> > > 1: volume management
> > > 2: type mgmt/glusterd
> > > 3: option rpc-auth.auth-glusterfs on
> > > 4: option rpc-auth.auth-unix on
> > > 5: option rpc-auth.auth-null on
> > > 6: option rpc-auth-allow-insecure on
> > > 7: option transport.socket.listen-backlog 128
> > > 8: option ping-timeout 30
> > > 9: option transport.socket.read-fail-log off
> > > 10: option transport.socket.keepalive-interval 2
> > > 11: option transport.socket.keepalive-time 10
> > > 12: option transport-type rdma
> > > 13: option working-directory /var/lib/glusterd
> > > 14: end-volume
> > > 15:
> > >
> >
> +------------------------------------------------------------------------------+
> > > [2016-01-29 05:03:56.002570] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 2
> > > [2016-01-29 05:03:56.003098] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 1
> > > [2016-01-29 05:03:56.003158] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 2
> > > [2016-01-29 05:03:56.855628] I [MSGID: 106493]
> > > [glusterd-rpc-ops.c:480:__glusterd_friend_add_cbk] 0-glusterd:
> > Received
> > > ACC from uuid: 388a8bb4-c530-44ff-838b-8f7b9e4c95db, host:
> 10.1.1.10,
> > > port: 0
> > > [2016-01-29 05:03:56.856787] I
> > [rpc-clnt.c:984:rpc_clnt_connection_init]
> > > 0-nfs: setting frame-timeout to 600
> > > [2016-01-29 05:03:57.859093] I [MSGID: 106540]
> > > [glusterd-utils.c:4191:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered MOUNTV3 successfully
> > > [2016-01-29 05:03:57.860228] I [MSGID: 106540]
> > > [glusterd-utils.c:4200:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered MOUNTV1 successfully
> > > [2016-01-29 05:03:57.861329] I [MSGID: 106540]
> > > [glusterd-utils.c:4209:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered NFSV3 successfully
> > > [2016-01-29 05:03:57.862421] I [MSGID: 106540]
> > > [glusterd-utils.c:4218:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered NLM v4 successfully
> > > [2016-01-29 05:03:57.863510] I [MSGID: 106540]
> > > [glusterd-utils.c:4227:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered NLM v1 successfully
> > > [2016-01-29 05:03:57.864600] I [MSGID: 106540]
> > > [glusterd-utils.c:4236:glusterd_nfs_pmap_deregister] 0-glusterd:
> > > De-registered ACL v3 successfully
> > > [2016-01-29 05:03:57.870948] W [socket.c:3009:socket_connect]
> 0-nfs:
> > > Ignore failed connection attempt on , (No such file or directory)
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > [2016-01-29 05:03:47.075614] W [socket.c:588:__socket_rwv]
> > > 0-data02-client-1: readv on 10.1.1.10:49155 <
> http://10.1.1.10:49155>
> > > <http://10.1.1.10:49155/> failed (No data available)
> > > [2016-01-29 05:03:47.076871] I [MSGID: 114018]
> > > [client.c:2042:client_rpc_notify] 0-data02-client-1: disconnected
> from
> > > data02-client-1. Client process will keep trying to connect to
> glusterd
> > > until brick's port is available
> > > [2016-01-29 05:03:47.170284] W [socket.c:588:__socket_rwv]
> 0-glusterfs:
> > > readv on 127.0.0.1:24007 <http://127.0.0.1:24007>
> > <http://127.0.0.1:24007/> failed (No data
> > > available)
> > > [2016-01-29 05:03:47.639163] W [socket.c:588:__socket_rwv]
> > > 0-data02-client-0: readv on 10.1.1.11:49153 <
> http://10.1.1.11:49153>
> > > <http://10.1.1.11:49153/> failed (No data available)
> > > [2016-01-29 05:03:47.639206] I [MSGID: 114018]
> > > [client.c:2042:client_rpc_notify] 0-data02-client-0: disconnected
> from
> > > data02-client-0. Client process will keep trying to connect to
> > glusterd
> > > until brick's port is available
> > > [2016-01-29 05:03:47.640222] E [MSGID: 108006]
> > > [afr-common.c:3880:afr_notify] 0-data02-replicate-0: All
> > subvolumes are
> > > down. Going offline until atleast one of them comes back up.
> > > [2016-01-29 05:03:57.872983] W [glusterfsd.c:1236:cleanup_and_exit]
> > > (-->/lib64/libpthread.so.0() [0x3e47a079d1]
> > > -->/usr/sbin/glusterfs(glusterfs_sigwaiter+0xcd) [0x405e6d]
> > > -->/usr/sbin/glusterfs(cleanup_and_exit+0x65) [0x4059d5] ) 0-: rec
> > > eived signum (15), shutting down
> > > [2016-01-29 05:03:58.881541] I [MSGID: 100030]
> > [glusterfsd.c:2318:main]
> > > 0-/usr/sbin/glusterfs: Started running /usr/sbin/glusterfs version
> > 3.7.6
> > > (args: /usr/sbin/glusterfs -s localhost --volfile-id
> > gluster/glustershd
> > > -p /var/lib/glusterd/
> > > glustershd/run/glustershd.pid -l /var/log/glusterfs/glustershd.log
> -S
> > > /var/run/gluster/8d72de580ccac07d2ecfc2491a9b1648.socket
> > --xlator-option
> > > *replicate*.node-uuid=9b103ea8-d248-44fc-8f80-3e87f7c4971c)
> > > [2016-01-29 05:03:58.890833] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 1
> > > [2016-01-29 05:03:59.340030] I [graph.c:269:gf_add_cmdline_options]
> > > 0-data02-replicate-0: adding option 'node-uuid' for volume
> > > 'data02-replicate-0' with value
> '9b103ea8-d248-44fc-8f80-3e87f7c4971c'
> > > [2016-01-29 05:03:59.342682] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 2
> > > [2016-01-29 05:03:59.342742] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 3
> > > [2016-01-29 05:03:59.342827] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 4
> > > [2016-01-29 05:03:59.342892] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 5
> > > [2016-01-29 05:03:59.342917] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 6
> > > [2016-01-29 05:03:59.343563] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 8
> > > [2016-01-29 05:03:59.343569] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 7
> > > [2016-01-29 05:03:59.343657] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 9
> > > [2016-01-29 05:03:59.343705] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 11
> > > [2016-01-29 05:03:59.343710] I [MSGID: 101190]
> > > [event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started
> > thread
> > > with index 10
> > > [2016-01-29 05:03:59.344278] I [MSGID: 114020]
> [client.c:2118:notify]
> > > 0-data02-client-0: parent translators are ready, attempting
> connect on
> > > transport
> > > [2016-01-29 05:03:59.346553] I [MSGID: 114020]
> [client.c:2118:notify]
> > > 0-data02-client-1: parent translators are ready, attempting
> connect on
> > > transport
> > > Final graph:
> > >
> >
> +------------------------------------------------------------------------------+
> > > 1: volume data02-client-0
> > > 2: type protocol/client
> > > 3: option ping-timeout 42
> > > 4: option remote-host gluster-stor01
> > > 5: option remote-subvolume /export/data/brick02
> > > 6: option transport-type socket
> > > 7: option username 5cc4f5d1-bcc8-4e06-ac74-520b20e2b452
> > > 8: option password 66b85782-5833-4f2d-ad0e-8de75247b094F
> > > 9: option event-threads 11
> > > 10: end-volume
> > > 11:
> > > 12: volume data02-client-1
> > > 13: type protocol/client
> > > 14: option ping-timeout 42
> > > 15: option remote-host gluster-stor02
> > > 16: option remote-subvolume /export/data/brick02
> > > 17: option transport-type socket
> > > 18: option username 5cc4f5d1-bcc8-4e06-ac74-520b20e2b452
> > > 19: option password 66b85782-5833-4f2d-ad0e-8de75247b094
> > > 20: option event-threads 11
> > > 21: end-volume
> > > 22:
> > > 23: volume data02-replicate-0
> > > 24: type cluster/replicate
> > > 25: option node-uuid 9b103ea8-d248-44fc-8f80-3e87f7c4971c
> > > 26: option background-self-heal-count 0
> > > 27: option metadata-self-heal on
> > > 28: option data-self-heal on
> > > 29: option entry-self-heal on
> > > 30: option self-heal-daemon enable
> > > 31: option iam-self-heal-daemon yes
> > > 32: subvolumes data02-client-0 data02-client-1
> > > 33: end-volume
> > > 34:
> > > 35: volume glustershd
> > > 36: type debug/io-stats
> > > 37: subvolumes data02-replicate-0
> > > 38: end-volume
> > > 39:
> > >
> >
> +------------------------------------------------------------------------------+
> > > [2016-01-29 05:03:59.348913] E [MSGID: 114058]
> > > [client-handshake.c:1524:client_query_portmap_cbk]
> 0-data02-client-1:
> > > failed to get the port number for remote subvolume. Please run
> > 'gluster
> > > volume status' on server to see if brick process
> > > is running.
> > > [2016-01-29 05:03:59.348960] I [MSGID: 114018]
> > > [client.c:2042:client_rpc_notify] 0-data02-client-1: disconnected
> from
> > > data02-client-1. Client process will keep trying to connect to
> > glusterd
> > > until brick's port is available
> > > [2016-01-29 05:03:59.436909] E [MSGID: 114058]
> > > [client-handshake.c:1524:client_query_portmap_cbk]
> 0-data02-client-0:
> > > failed to get the port number for remote subvolume. Please run
> > 'gluster
> > > volume status' on server to see if brick process
> > > is running.
> > > [2016-01-29 05:03:59.436974] I [MSGID: 114018]
> > > [client.c:2042:client_rpc_notify] 0-data02-client-0: disconnected
> from
> > > data02-client-0. Client process will keep trying to connect to
> > glusterd
> > > until brick's port is available
> > > [2016-01-29 05:03:59.436991] E [MSGID: 108006]
> > > [afr-common.c:3880:afr_notify] 0-data02-replicate-0: All
> > subvolumes are
> > > down. Going offline until atleast one of them comes back up.
> > > [2016-01-29 05:04:02.886317] I [rpc-clnt.c:1847:rpc_clnt_reconfig]
> > > 0-data02-client-0: changing port to 49153 (from 0)
> > > [2016-01-29 05:04:02.888761] I [rpc-clnt.c:1847:rpc_clnt_reconfig]
> > > 0-data02-client-1: changing port to 49155 (from 0)
> > > [2016-01-29 05:04:02.891105] I [MSGID: 114057]
> > > [client-handshake.c:1437:select_server_supported_programs]
> > > 0-data02-client-0: Using Program GlusterFS 3.3, Num (1298437),
> > Version (330)
> > > [2016-01-29 05:04:02.891360] I [MSGID: 114046]
> > > [client-handshake.c:1213:client_setvolume_cbk] 0-data02-client-0:
> > > Connected to data02-client-0, attached to remote volume
> > > '/export/data/brick02'.
> > > [2016-01-29 05:04:02.891373] I [MSGID: 114047]
> > > [client-handshake.c:1224:client_setvolume_cbk] 0-data02-client-0:
> > Server
> > > and Client lk-version numbers are not same, reopening the fds
> > > [2016-01-29 05:04:02.891403] I [MSGID: 108005]
> > > [afr-common.c:3841:afr_notify] 0-data02-replicate-0: Subvolume
> > > 'data02-client-0' came back up; going online.
> > > [2016-01-29 05:04:02.891518] I [MSGID: 114035]
> > > [client-handshake.c:193:client_set_lk_version_cbk]
> 0-data02-client-0:
> > > Server lk version = 1
> > > [2016-01-29 05:04:02.893074] I [MSGID: 114057]
> > > [client-handshake.c:1437:select_server_supported_programs]
> > > 0-data02-client-1: Using Program GlusterFS 3.3, Num (1298437),
> > Version (330)
> > > [2016-01-29 05:04:02.893251] I [MSGID: 114046]
> > > [client-handshake.c:1213:client_setvolume_cbk] 0-data02-client-1:
> > > Connected to data02-client-1, attached to remote volume
> > > '/export/data/brick02'.
> > > [2016-01-29 05:04:02.893276] I [MSGID: 114047]
> > > [client-handshake.c:1224:client_setvolume_cbk] 0-data02-client-1:
> > Server
> > > and Client lk-version numbers are not same, reopening the fds
> > > [2016-01-29 05:04:02.893401] I [MSGID: 114035]
> > > [client-handshake.c:193:client_set_lk_version_cbk]
> 0-data02-client-1:
> > > Server lk version = 1
> > >
> > >
> > > _______________________________________________
> > > Gluster-devel mailing list
> > > Gluster-devel at gluster.org <mailto:Gluster-devel at gluster.org>
> > > http://www.gluster.org/mailman/listinfo/gluster-devel
> > >
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-devel/attachments/20160202/dd1a8701/attachment-0001.html>
More information about the Gluster-devel
mailing list