Summary
OpsBridge Containerized - Unable to start uCMDB pod after side-by-side upgrade 2019.11 -> 2020.10 with an external Postgres/external Amazon Postgres RDS.
Error
After performing a side-by-side upgrade of containerized OpsBridge 2019.11 to 2020.10 with an external Amazon Postgres RDS( Relational Database Service ) the itom-ucmdb pod fails to start and the following error is observed in the pod's log and ucmdb log:
The statement “INSERT INTO CONTENT_PACKS (VERSION,CP_BYTES) values (?,?); Values: '2020.11.100',java.io.FileInputStream@50f02f86
caused by cancelling statement due to statement timeout
Cause
in Amazon Postgres RDS( Relational Database Service ) DB, statement_timeout is set to 10s.
any query which take longer than 10sec would close the connection.
Fix
If you see the error as "cancelling statement due to statement timeout"
check the statement_timeout after log in to respective DB by querying "show statement_timeout".
The recommendation is to switch it off.
Similar to the postgresql official documentation we would not recommend setting it globally which also affects rtsm, event, mgmt databases of OBM: “Setting statement_timeout in postgresql.conf is not recommended because it would affect all sessions.” (https://www.postgresql.org/docs/13/runtime-config-client.html)