Major-Upgrades

CPO enables the use of the in-place upgrade, which makes it possible to upgrade a cluster to a new PG major. For this purpose, pg_upgrade is used in the background.

PAY ATTENTION: Note that an in-place upgrade generates both a pod restore in the form of a rolling update and an operational interruption of the cluster during the actual execution of the restore.

1. Pod restart - Use the rolling update strategy to replace all pods based on the new ENV PGVERSION with the version you want to update to.
2. Check - Check that the new PGVERSION is larger than the previously used one.
3. Check whether the new PGVERSION is larger than the previously used one and the maintenance mode of the cluster must be deactivated. In addition, the replicas should not have a high lag.

1. use initdb to prepare a new data_dir (data_new) based on the new PGVERSION.
2. check the upgrade possibility with pg_upgrade --check

HINT: If one of the steps is aborted, a cleanup is performed

1. remove dependencies that can cause problems. For example, the extensions pg_stat_statements and pgaudit.
2. activate the maintenance mode of the cluster
3. terminate PostgreSQL in an orderly manner
4. check pg_controldata for the checkpoint position and wait until all replicas apply the latest checkpoint location
5. use port 5432 for rsyncd and start it

1. Call pg_upgrade -k to start the Upgrade

ATTENTION if the process failed, we need to rollback, if it was sucessful we’re reaching the point of no return

2. Rename the directories. data -> data_old and data_new -> data
3. Update the Patroni.config (postgres.yml)
4. Call Checkpoint on every replica and trigger rsync on the Replicas
5. Wait for Replicas to complete rsxnc. Timeout: 300
6. Stop rsyncd on Primary and remove ininitialize key from DCS, because its based on the old sysid
7. Start Patroni on the Primary and start the postgres locally
8. Reset custom staticstics, warmup the Memory and start Analyze in stages in separate threads
9. Wait for every Replica to become ready
10. Disable the maintenance mode for the Cluster
11. Restore custom statistics, analyze these tables and restore dropped objetcs from Prepare the upgrade

1. Drop directory data_old
2. Trigger new Backup

1. Stop rsynd if its running
2. Disable the maintenance mode for the Cluster
3. Drop directory data_new

spec:
  postgresql:
    version: "16"

To trigger an In-Place-Upgrade you have just to increase the parameter spec.postgresql.version. If you choose a valid number the Operator will start with the prozedure, described above. If you choosse a not allowed value, you will give an error and if you decrease the value, the operator will just ignore it with the following log-Entry.

Operator-Log