Hi,

We have a managed postgres database with digital ocean with a large number of tables. Running a normal pg_dump will crash as it needs a lock per table and is exceeding whatever the managed clusters configured postgresql max_locks_per_transaction is, which I do not believe we can change ourselves.

pg_dump: WARNING:  out of shared memory
pg_dump: SQL command failed
pg_dump: Error message from server: ERROR:  out of shared memory
HINT:  You might need to increase max_locks_per_transaction.

We cannot instead use pg_basebackup to do backups as the configured database user/host does not have permissions to connect as a replication connection.

And so is it literally impossible via normal postgres commands to backup an entire digital ocean managed database if you go over a certain table limit?

Thanks