How to enable streaming replication if we clone database disks to new virtual machine?

Question

On PostgreSQL v15 on RedHat 9 we would like to establish asynchronous read-only standby streaming replication.

Primary server has 10 TB of data and pretty much a lot of new transactions, because it is big data server.

I have seen several instructions how to establish standby replication, but all of the instructions have part to do it with "pg_basebackup" utility that requires to have empty directory on standby server. This solution does not fit our needs because we can't afford to copy data using backup utility, because it just takes way too long. We want to clone database disks with disk storage facility that is super fast.

What we would like to do is:

1. From primary server clone the disks where postgres cluster is located using disk storage facility, which is super fast.
2. Attach cloned disks to new Linux vmware virtual machine (standby server). Start standby server. I can confirm postgresql database is working fine on standby server, but it is currently not replicating any data from primary server.
3. Setup replication, so server starts replicating data to standby server. --> This is the step I am struggling now.

How to establish streaming replication using cloned disks?

Answer 1

Instead of running pg_basebackup, you can use the low-level backup API to perform a physical backup that can be used to initialize a standby server.

See step 3 of the instructions:

3. Perform the backup, using any convenient file-system-backup tool such as tar or cpio (not pg_dump or pg_dumpall). It is neither necessary nor desirable to stop normal operation of the database while you do this. See Section 26.3.3.1 for things to consider during this backup.

You can perform that step by cloning your disks.