[ https://issues.apache.org/jira/browse/IGNITE-9793?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Pavel Voronkin updated IGNITE-9793: ----------------------------------- Attachment: 724_gc_server_memory.log.0.current.7z > Deactivation, segmentation of one node, activation may lead to hang > activation forever > -------------------------------------------------------------------------------------- > > Key: IGNITE-9793 > URL: https://issues.apache.org/jira/browse/IGNITE-9793 > Project: Ignite > Issue Type: Bug > Affects Versions: 2.5 > Reporter: Pavel Voronkin > Priority: Major > Attachments: 601_gc_server_memory.log.0.current.7z, > 724_gc_server_memory.log.0.current.7z > > > There is topology with ring of nodes: > *665(coordinator) > 601 > 724 > 910 > 655 > ...* deactivated > *665(coordinator)* *> 601 > {color:#FF0000}724{color} > 910 > 655 > ...* > ****node failed > *665(coordinator) > 601 > 910 > 655 > ...* **activated > During activation node 910 haven't received StateChangedMessage, hovever 655 > and all subsequent nodes received and responded to coordinator. > So coordinator expects to have 154 messages but received only 153 that is why > activation hangs. > Details bellow: > > Coordinator deactivated: > 2018-09-24 15:09:01.609 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Successfully deactivated data structures, services and caches > [nodeId=e002e011-8d1c-4353-a0f3-b71264c5b0f4, client=false, > topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]] > 2018-09-24 15:09:01.620 > [DEBUG][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.l.ExchangeLatchManager] > Server latch is created [latch=CompletableLatchUid > {id='exchange', topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]} > , participantsSize=160] > 2018-09-24 15:09:01.621 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%] > nodes 601, 724, 910, 655 were deactivated: > 2018-09-24 15:09:01.609 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Successfully deactivated data structures, services and caches > [nodeId=e002e011-8d1c-4353-a0f3-b71264c5b0f4, client=false, > topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]] > 2018-09-24 15:09:03.328 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Successfully deactivated data structures, services and caches > [nodeId=22a58223-47b5-43c2-897b-e70e8e50edf7, client=false, > topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]] > 2018-09-24 15:09:03.334 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Successfully deactivated data structures, services and caches > [nodeId=973eb8ce-3b8c-463d-a6ab-00ac66d93f13, client=false, > topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]] > 2018-09-24 15:09:03.332 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Successfully deactivated data structures, services and caches > [nodeId=a904bac4-aaed-4f69-90f3-c13bc4d331d1, client=false, > topVer=AffinityTopologyVersion [topVer=183, minorTopVer=1]] > Spi on coordinator received node 724 failed message: > 2018-09-24 15:17:00.220 [WARN > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > Node FAILED: TcpDiscoveryNode [id=a904bac4-aaed-4f69-90f3-c13bc4d331d1, > addrs=ArrayList [10.116.206.98], sockAddrs=HashSet > [grid724.domain/10.116.206.98:47500], discPort=47500, order=110, > intOrder=110, lastExchangeTime=1537528210290, loc=false, > ver=2.5.1#20180906-sha1:ebde6c79, isClient=false] > 2018-09-24 15:17:00.221 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > Topology snapshot [ver=187, servers=156, clients=0, CPUs=8736, > offheap=190000.0GB, heap=4800.0GB] > 2018-09-24 15:17:00.221 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > ^-- Node [id=E002E011-8D1C-4353-A0F3-B71264C5B0F4, clusterState=INACTIVE] > 2018-09-24 15:17:00.221 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > ^-- Baseline [id=6, size=160, online=156, offline=4] > topology rolled to version 187, then another node 931 failed: > 2018-09-24 15:17:00.466 [WARN > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > Node FAILED: TcpDiscoveryNode [id=83536b6d-8aa3-4c85-b3da-5e577ae37ac6, > addrs=ArrayList [10.116.215.3], sockAddrs=HashSet > [grid931.ca.sbrf.ru/10.116.215.3:47500], discPort=47500, order=73, > intOrder=73, lastExchangeTime=1537528186599, loc=false, > ver=2.5.1#20180906-sha1:ebde6c79, isClient=false] > 2018-09-24 15:17:00.467 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > Topology snapshot [ver=188, servers=155, clients=0, CPUs=8680, > offheap=190000.0GB, heap=4800.0GB] > 2018-09-24 15:17:00.467 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > ^-- Node [id=E002E011-8D1C-4353-A0F3-B71264C5B0F4, clusterState=INACTIVE] > 2018-09-24 15:17:00.467 [INFO > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > ^-- Baseline [id=6, size=160, online=155, offline=5] > topology rolled to version 188. > Coordinator rejects node 724 form joining the cluster: > 2018-09-24 15:17:27.491 [INFO > ][grid-nio-worker-tcp-comm-9-#129%DPL_GRID%DplGridNodeName%|#129%DPL_GRID%DplGridNodeName%][o.a.i.s.c.tcp.TcpCommunicationSpi] > Accepted incoming communication connection [locAddr=/10.116.206.11:47100, > rmtAddr=/10.116.206.98:41611] > 2018-09-24 15:17:27.491 [WARN > ][grid-nio-worker-tcp-comm-9-#129%DPL_GRID%DplGridNodeName%|#129%DPL_GRID%DplGridNodeName%][o.a.i.s.c.tcp.TcpCommunicationSpi] > Close incoming connection, unknown node > [nodeId=a904bac4-aaed-4f69-90f3-c13bc4d331d1, ses=GridSelectorNioSessionImpl > [worker=GridWorker [name=grid-nio-worker-tcp-comm-9, > igniteInstanceName=DPL_GRID%DplGridNodeName, finished=false, > hashCode=632168435, interrupted=false, > runner=grid-nio-worker-tcp-comm-9-#129%DPL_GRID%DplGridNodeName%|#129%DPL_GRID%DplGridNodeName%]AbstractNioClientWorker > [idx=9, bytesRcvd=326476100, bytesSent=39876884360, bytesRcvd0=42, > bytesSent0=18, select=true, super=]DirectNioClientWorker [super=], > writeBuf=java.nio.DirectByteBuffer[pos=0 lim=32768 cap=32768], > readBuf=java.nio.DirectByteBuffer[pos=42 lim=42 cap=32768], inRecovery=null, > outRecovery=null, super=GridNioSessionImpl [locAddr=/10.116.206.11:47100, > rmtAddr=/10.116.206.98:41611, createTime=1537791447487, closeTime=0, > bytesSent=18, bytesRcvd=42, bytesSent0=18, bytesRcvd0=42, > sndSchedTime=1537791447487, lastSndTime=1537791447487, > lastRcvTime=1537791447487, readsPaused=false, > filterChain=FilterChain[filters=[, GridConnectionBytesVerifyFilter], > accepted=true, markedForClose=false]]] > > Node 724 SEGMENTED before activation starts: > 2018-09-24 15:17:50.068 [WARN > ][tcp-disco-msg-worker-#2%DPL_GRID%DplGridNodeName%|#2%DPL_GRID%DplGridNodeName%][o.a.i.s.d.tcp.TcpDiscoverySpi] > Node is out of topology (probably, due to short-time network problems). > 2018-09-24 15:17:50.069 [WARN > ][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][o.a.i.i.m.d.GridDiscoveryManager] > Local node SEGMENTED: TcpDiscoveryNode > [id=a904bac4-aaed-4f69-90f3-c13bc4d331d1, addrs=ArrayList [10.116.206.98], > sockAddrs=HashSet [grid724.domain/10.116.206.98:47500], discPort=47500, > order=110, intOrder=110, lastExchangeTime=1537791470063, loc=true, > ver=2.5.1#20180906-sha1:ebde6c79, isClient=false] > 2018-09-24 15:18:07.145 > [ERROR][disco-event-worker-#152%DPL_GRID%DplGridNodeName%|#152%DPL_GRID%DplGridNodeName%][org.apache.ignite.Ignite] > Critical system error detected. Will be handled accordingly to configured > handler [hnd=class o.a.i.failure.StopNodeFailureHandler, > failureCtx=FailureContext [type=SEGMENTATION, err=null]] > Coordinator started activating topology without nodes 724 and 931. > 2018-09-24 15:19:48.686 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Start activation process [nodeId=e002e011-8d1c-4353-a0f3-b71264c5b0f4, > client=false, topVer=AffinityTopologyVersion [topVer=188, minorTopVer=1]] > But node 910 which is next to node 724 haven't received activation message. > Hovever 655 which is next after 910 received ChangeStateMessage: > 2018-09-24 15:21:08.430 [INFO > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Start activation process [nodeId=22a58223-47b5-43c2-897b-e70e8e50edf7, > client=false, topVer=AffinityTopologyVersion [topVer=188, minorTopVer=1]] > Coordinator received all expected messages except 910, that was: > 2018-09-24 15:24:25.911 [INFO > ][sys-#28144%DPL_GRID%DplGridNodeName%|#28144%DPL_GRID%DplGridNodeName%][o.a.i.i.p.c.d.d.p.GridDhtPartitionsExchangeFuture] > Coordinator received single message [ver=AffinityTopologyVersion > [topVer=188, minorTopVer=1], node=073f1598-6b70-49df-8f45-126735611775, > *allReceived=false*] > > GridDhtPartitionsExchangeFuture hangs forever: > 09-24 15:33:46.784 [WARN > ][exchange-worker-#153%DPL_GRID%DplGridNodeName%|#153%DPL_GRID%DplGridNodeName%][o.apache.ignite.internal.diagnostic] > Failed to wait for partition map exchange [topVer=AffinityTopologyVersion > [topVer=188, minorTopVer=1], node=e002e011-8d1c-4353-a0f3-b71264c5b0f4]. > Dumping pending objects that might be the cause: > > So one node in the ring missed the message, hovever all other nodes in > topology got it, how is that possible? > -- This message was sent by Atlassian JIRA (v7.6.3#76005)