Symptoms
- The server OS was updated by using dist-upgrade
During an attempt to connect the PostgreSQL instance via Tools & Settings > Database Servers with hostname localhost and any login and password, the process fails and Plesk produces an Internal Server error
PostgreSQL variables in the
/etc/psa/psa.confare malformed and are similar to the following:PGSQL_DATA_D /var/lib/postgresql/11
16/main
main
PGSQL_CONF_D /etc/postgresql/11
16/main
main
PGSQL_BIN_D /usr/lib/postgresql/11
16/binThe configuration file
/etc/psa/psa.confgot malformed further every time/opt/psa/admin/bin/pg_manage --statusranThere are no running PostgreSQL instances:
# /usr/bin/pg_lsclusters -h
11 main 5432 down postgres /var/lib/postgresql/11/main /var/log/postgresql/postgresql-11-main.log
16 main 5432 online postgres /var/lib/postgresql/16/main /var/log/postgresql/postgresql-16-main.log
Cause
This behavior is tied to a known bug with ID PPPM-13771, which will be resolved in future versions of Plesk.
You may track it in the Change Log for Plesk Obsidian.
Resolution
Until the bug is fixed permanently, you may apply the following workaround:
Workaround
Connect to the server via SSH
Check which PostgreSQL instance is down
# /usr/bin/pg_lsclusters -h
11 main 5432 down postgres /var/lib/postgresql/11/main /var/log/postgresql/postgresql-11-main.log
16 main 5432 online postgres /var/lib/postgresql/16/main /var/log/postgresql/postgresql-16-main.logNote: In this case, the instance 11 is down.
Create a backup folder to store the instance that is down
# mkdir ~/pgsql_etc_bkp
Move the instance that is down to the backup folder
# mv /etc/postgresql/11 ~/pgsql_etc_bkp/
Correct the PostgreSQL configuration in
/etc/psa/psa.confby using the following example:# PostgreSQL
PGSQL_DATA_D /var/lib/postgresql/16/main
PGSQL_CONF_D /etc/postgresql/16/main
PGSQL_BIN_D /usr/lib/postgresql/16/binNote: Where 16 is the online instance in this case.
Reread the PostgreSQL configuration by executing the following command:
# plesk sbin pg_manage --check-installed
