Details
-
Improvement
-
Resolution: Won't Do
-
Minor - P4
-
None
Description
Description
The cursor.maxTimeMS Behavior section doesn't explain what is reported in the logs.
At least in 3.6.12, there will be a log entry with errmsg: "operation exceeded time limit" such as:
2019-10-15T20:02:11.341+0100 I COMMAND [conn636131] command myDb.myColl command: aggregate { aggregate: "myColl", pipeline: [... ], allowDiskUse: true, fromMongos: true, cursor: { batchSize: 101 }, maxTimeMS: 5000, shardVersion: [ Timestamp(19622, 240811), ObjectId('5705031ca624b38123a80989') ], $clusterTime: { clusterTime: Timestamp(0, 0), signature: { hash: BinData(0, 0000000000000000000000000000000000000000), keyId: 0 } }, ... $db: "myDb" } planSummary: IXSCAN { ... } exception: Error in $cursor stage :: caused by :: errmsg: "operation exceeded time limit" code:ExceededTimeLimit numYields:241 reslen:276 locks:{ Global: { acquireCount: { r: 586 } }, Database: { acquireCount: { r: 293 } }, Collection: { acquireCount: { r: 293 } } } protocol:op_msg 5011ms
|
It's useful to know that this is logged with this message so we can search for this type of slow query.