create a runbook for interact with some use cases of data corruption

We need to have at least an initial guide on how to react in case of data corruption.

I would consider the following initial steps:

• How to verify if we have more blocks corrupted in the database.
• How to identify what is the database object that is corrupted.
• Some initial troubleshooting in some cases.

In case of a secondary node:

• drain the traffic and recreate a new node.

In the case of the primary:

• Verify if it is an index:

  -If yes recreate CONCURRENTLY the index.

If it is a table verify what rows are affected.

Please contact support ( DBRE and Ongres As soon as Possible)