Details
-
Bug
-
Status: Closed
-
Major - P3
-
Resolution: Duplicate
-
None
-
None
-
None
-
ALL
-
-
0
Description
If you bring up a sharded cluster on 3.7.5 with a mmapv1 multi-node shard, you get a timeout error from trying to add the shard on the mongos:
mongos> db.runCommand({"addShard" : "a/louisamac:9001,louisamac:9002,louisamac:9003", name: "shard0"})
|
{
|
"ok" : 0,
|
"errmsg" : "failed to run command { setFeatureCompatibilityVersion: \"4.0\" } when attempting to add shard a/louisamac:9001,louisamac:9002,louisamac:9003 :: caused by :: NetworkInterfaceExceededTimeLimit: timed out",
|
"code" : 96,
|
"codeName" : "OperationFailed",
|
"$clusterTime" : {
|
"clusterTime" : Timestamp(1524674234, 1),
|
"signature" : {
|
"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
|
"keyId" : NumberLong(0)
|
}
|
},
|
"operationTime" : Timestamp(1524674234, 1)
|
}
|
Note: I tried this on WT and it worked fine. Assigning to storage for that reason, but may need to be reassigned.
Second note: Also did not have this problem with a 1-node shard.
Attached config files and log files.
Attachments
Issue Links
- duplicates
-
SERVER-34483 Avoid taking DBLocks when clearing in-memory database cache in FCV op observer if possible
-
- Closed
-
- is related to
-
SERVER-27534 All writing operations must fail if the term changes
-
- Closed
-
-
SERVER-34459 Clear in-memory database versions on setFCV downgrade on shard primaries and secondaries
-
- Closed
-