Assertion "can't find shard for" when dropping sharded collection

XMLWordPrintableJSON

    • Type: Bug
    • Resolution: Done
    • Priority: Major - P3
    • None
    • Affects Version/s: 2.4.8
    • Component/s: Sharding
    • None
    • Environment:
      Debian Squeeze
      MongoS 2.4.8
      PyMongo 2.0.1
    • Sharding
    • ALL
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      MongoS can come in a state where dropping a sharded collection fails. Namespaces are no longer existent on shards, but configsvr entry in "collections" collection is still

      {"dropped" : false}

      After restart of MongoS dropping works fine.

      Log Message in MongoS:

      Fri Jan 17 09:12:32.201 [conn59971] DROP: database001.collection001
      Fri Jan 17 09:12:32.201 [conn59971] about to log metadata event: { _id: "mongos001-2014-01-17T09:12:32-52d8f40058a9bbeefff45009", server: "mongos001", clientAddr: "N/A", time: new Date(1389949952201), what: "dropCollectio
      n.start", ns: "database001.collection001", details: {} }
      Fri Jan 17 09:12:32.680 [conn59971] distributed lock 'database001.collection001/mongos001:27017:1386839918:1804289383' acquired, ts : 52d8f40058a9bbeefff4500a
      Fri Jan 17 09:12:33.420 [conn59971] Assertion: 13129:can't find shard for: shard001/server002:27018,server001:27018
      0xa89e91 0xa4facb 0xa5000c 0x9a3568 0x99e874 0x8c856f 0x8c9e6b 0x9204b7 0x9994c6 0x90eb1b 0x9cae73 0x99716d 0x6666c4 0xa7645e 0x7fcd1eb598ca 0x7fcd1df0cb6d
       /usr/bin/mongos(_ZN5mongo15printStackTraceERSo+0x21) [0xa89e91]
       /usr/bin/mongos(_ZN5mongo11msgassertedEiPKc+0x9b) [0xa4facb]
       /usr/bin/mongos() [0xa5000c]
       /usr/bin/mongos(_ZN5mongo15StaticShardInfo4findERKSs+0x358) [0x9a3568]
       /usr/bin/mongos(_ZN5mongo5Shard5resetERKSs+0x34) [0x99e874]
       /usr/bin/mongos(_ZN5mongo15setShardVersionERNS_12DBClientBaseERKSsNS_12ChunkVersionEN5boost10shared_ptrIKNS_12ChunkManagerEEEbRNS_7BSONObjE+0x30f) [0x8c856f]
       /usr/bin/mongos(_ZNK5mongo12ChunkManager4dropEN5boost10shared_ptrIKS0_EE+0x10db) [0x8c9e6b]
       /usr/bin/mongos(_ZN5mongo15dbgrid_pub_cmds7DropCmd3runERKSsRNS_7BSONObjEiRSsRNS_14BSONObjBuilderEb+0x267) [0x9204b7]
       /usr/bin/mongos(_ZN5mongo7Command22execCommandClientBasicEPS0_RNS_11ClientBasicEiPKcRNS_7BSONObjERNS_14BSONObjBuilderEb+0x416) [0x9994c6]
       /usr/bin/mongos(_ZN5mongo7Command20runAgainstRegisteredEPKcRNS_7BSONObjERNS_14BSONObjBuilderEi+0x21b) [0x90eb1b]
       /usr/bin/mongos(_ZN5mongo14SingleStrategy7queryOpERNS_7RequestE+0x2d3) [0x9cae73]
       /usr/bin/mongos(_ZN5mongo7Request7processEi+0x1bd) [0x99716d]
       /usr/bin/mongos(_ZN5mongo21ShardedMessageHandler7processERNS_7MessageEPNS_21AbstractMessagingPortEPNS_9LastErrorE+0x74) [0x6666c4]
       /usr/bin/mongos(_ZN5mongo17PortMessageServer17handleIncomingMsgEPv+0x42e) [0xa7645e]
       /lib/libpthread.so.0(+0x68ca) [0x7fcd1eb598ca]
       /lib/libc.so.6(clone+0x6d) [0x7fcd1df0cb6d]
      Fri Jan 17 09:12:33.423 [conn59971] scoped connection to shard001/server002:27018,server001:27018 not being returned to the pool
      Fri Jan 17 09:12:33.555 [conn59971] distributed lock 'database001.collection001/mongos001:27017:1386839918:1804289383' unlocked.
      Fri Jan 17 09:12:33.821 [conn59971] end connection 10.3.44.13:56460 (9 connections now open)
      

            Assignee:
            [DO NOT USE] Backlog - Sharding Team
            Reporter:
            Simon Schneider
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: