Details
-
Bug
-
Resolution: Fixed
-
Major
-
3.0.0-beta.4
-
1
Description
Seeing an error from the SDK under this Query scenario, with MH only:
Operations: Continuous n1ql queries
Change: Add a n1ql node in and remove another and rebalance
Cluster:
- kv
- n1ql <- not in the cluster at the start, then rebalanced in
- index
- n1ql <- rebalanced out
Error: All queries will timeout after the change. In the logs it looks like the error is because of QueryPreparedStatementFailure.
Logs: SdkdConsole.log.zip . and full run logs http://sdkbuilds.sc.couchbase.com/view/.NET/job/sdk-net-situational-release/job/SDK3.0-Release/job/net-centos-madhatter/7/console (not very informative in this case)
Note: There is also a failure with the service restart query also due to the QueryPreparedStatementFailure, as seen on the graph http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-6.5.0-4960/SvcRestartQuery-N1QL/04-01-20/042644/8ba514235eb2fc2b9342b79c90b95b1a-N1QL.html
Im attaching further logs for this test too, although I think you'll find they show the same issue SdkdConsole-SvcRst.log.zip
Attachments
Issue Links
- relates to
-
NCBC-2388 Ensure that prepared statements that are previously created are not retried
- Resolved