[SERVER-17624] Interrupting aggregation operation can trip fatal assertion Created: 17/Mar/15  Updated: 21/Mar/16  Resolved: 21/Mar/16

Status: Closed
Project: Core Server
Component/s: Aggregation Framework
Affects Version/s: 3.0.0
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: J Rassi Assignee: Ramon Fernandez Marina
Resolution: Cannot Reproduce Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
is related to SERVER-16607 Dropping collection during active agg... Closed
is related to SERVER-22541 Aggregation plan executors should be ... Closed
Operating System: ALL
Sprint: Quint Iteration 3.1.1
Participants:

 Description   

Interrupting an aggregation operation can trip invariant( _cursor ) in ClientCursorPin::deleteUnderlying().

Example assertion failure:

 mongod(_ZN5mongo15printStackTraceERSo+0x29) [0xf5b1f9]
 mongod(_ZN5mongo10logContextEPKc+0xE1) [0xefba11]
 mongod(_ZN5mongo15invariantFailedEPKcS1_j+0xB2) [0xee0ce2]
 mongod(_ZN5mongo15ClientCursorPin16deleteUnderlyingEv+0x72) [0x94ce52]
 mongod(_ZN5mongo15PipelineCommand3runEPNS_16OperationContextERKSsRNS_7BSONObjEiRSsRNS_14BSONObjBuilderEb+0xF2C) [0x9a27ec]
 mongod(_ZN5mongo12_execCommandEPNS_16OperationContextEPNS_7CommandERKSsRNS_7BSONObjEiRSsRNS_14BSONObjBuilderEb+0x34) [0x9f1744]
 mongod(_ZN5mongo7Command11execCommandEPNS_16OperationContextEPS0_iPKcRNS_7BSONObjERNS_14BSONObjBuilderEb+0xC13) [0x9f2683]
 mongod(_ZN5mongo12_runCommandsEPNS_16OperationContextEPKcRNS_7BSONObjERNS_11_BufBuilderINS_16TrivialAllocatorEEERNS_14BSONObjBuilderEbi+0x28B) [0x9f327b]
 mongod(_ZN5mongo8runQueryEPNS_16OperationContextERNS_7MessageERNS_12QueryMessageERKNS_15NamespaceStringERNS_5CurOpES3_b+0x755) [0xbb3d05]
 mongod(_ZN5mongo16assembleResponseEPNS_16OperationContextERNS_7MessageERNS_10DbResponseERKNS_11HostAndPortEb+0xB19) [0xacc1e9]
 mongod(_ZN5mongo16MyMessageHandler7processERNS_7MessageEPNS_21AbstractMessagingPortEPNS_9LastErrorE+0xE0) [0x828f70]
 mongod(_ZN5mongo17PortMessageServer17handleIncomingMsgEPv+0x32B) [0xf0f36b]
 libpthread.so.0(+0x7F18) [0x7f7c1a528f18]
 libc.so.6(clone+0x6D) [0x7f7c18fe9b9d]



 Comments   
Comment by J Rassi [ 21/Mar/16 ]

We have been unable to reproduce this issue and have seen no further reports, so I am closing this ticket as "Cannot Reproduce". In addition, it is likely that this issue is closely related to SERVER-22541, which is scheduled to will be fixed for the 3.4 or 3.6 release.

Please re-open this ticket if you have encountered this issue, and we will continue our investigation.

Comment by J Rassi [ 07/Apr/15 ]

The filing of this ticket was prompted by a report we received in which a 3.0.0 deployment tripped this assertion failure on mongod during a killOp against an aggregation operation. We do not yet have a reproducible case, and this ticket remains active.

Comment by David Murphy [ 07/Apr/15 ]

Is there any more movement on this, I would like to understand if this is a real issue we should expect with our clients.

Comment by J Rassi [ 17/Mar/15 ]

The assertion in question was introduced in SERVER-16607, and is triggered upon double-deletion of a ClientCursorPin. I suspect that this is a regression introduced by that ticket, but can't confirm yet. I'll update this ticket with more details once I am able to reproduce this issue.

Generated at Thu Feb 08 03:45:05 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.