Encountered issue after a power failure which cause all the nodes of the identity manager cluster to reboot.
Based on the error, i’m getting the following:
Error 500: org.hibernate.exception.GenericJDBCException: could not prepare statement
Bookmark this KB as this explains how the internal Postgres cluster works – basically, if all the nodes failed – you need to manully bring up the VIP of the postgres DB for it to work:
https://kb.vmware.com/s/article/75080
High level of what needs to happen:
- Verify if the pgIP VIP is present on the primary
- To determine the primary execute the following: su postgres -c “echo -e ‘password’|/opt/vmware/vpostgres/current/bin/psql -h localhost -p 9999 -U pgpool postgres -c \”show pool_nodes\””
- Login to the primary and recreate the VIP if not present
- Recover failed/ down nodes by doing the following: /usr/local/bin/pcp_recovery_node -h delegateIP -p 9898 -U pgpool -n <node_id>
- NOTE: delegateIP is the pg VIP.
Bonus link: shutdown/power-up sequence for IDM nodes
