-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: 3.6.0-rc0
-
Component/s: Querying, Replication
-
None
-
Fully Compatible
-
ALL
-
v3.6
-
- start a single node replica set
- cd into mongo repository
- download the attached repro script: change_streams_repro.js
- run the script:
> mongo change_streams_repro.js
-
Query 2017-11-13, Query 2017-12-04, Query 2017-12-18, Query 2018-01-01
-
0
-
None
-
None
-
None
-
None
-
None
-
None
-
None
On a 3.6.0-rc0 single node replica set, while applying a CRUD workload with 16 or more threads, a change stream using the option fullDocument: updateLookup eventually fails during a getMore with the following error:
2017-10-23T16:51:32.883-0400 I QUERY [thread2] js thread raised js exception: JSInterpreterFailure: Error: getMore command failed: {
"operationTime" : Timestamp(1508791892, 30247),
"ok" : 0,
"errmsg" : "collection or index disappeared when cursor yielded: ExceededTimeLimit: operation exceeded time limit",
"code" : 175,
"codeName" : "QueryPlanKilled",
"$clusterTime" : {
"clusterTime" : Timestamp(1508791892, 30249),
"signature" : {
"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
"keyId" : NumberLong(0)
}
}
}
- duplicates
-
SERVER-31597 Refactor $changeStream post-update lookup
-
- Closed
-