Skip to content

Conversation

alexadriaanse
Copy link

This change is to prevent scenarios where a long-running database recovery would cause pg_ctl to time out, causing Docker to kill the container before it had a chance to recover.

This change is to prevent scenarios where a long-running database recovery would cause pg_ctl to time out, causing Docker to kill the container before it had a chance to recover.
@tianon
Copy link
Member

tianon commented Dec 15, 2015

I'm not sure I fully understand what this change fixes -- database recovery wouldn't be happening in this block, right? (since this block is only for database initialization, so I'm confused how you'd be doing recovery during one of these pg_ctl invocations)

@alexadriaanse
Copy link
Author

Ah, you're right. I have a wrapper entrypoint that my company is using internally which adds some functionality to the standard entrypoint. This internal entrypoint was running into these pg_ctl timeout issues. I fixed them in a number of places, including the standard docker-entrypoint.sh, but had forgotten that pg_ctl is not called by docker-entrypoint.sh after the database is initialized and therefore these changes are not necessary here.

Sorry for the false alarm.

@alexadriaanse alexadriaanse deleted the prevent-pgctl-timeout branch December 16, 2015 20:01
@tianon
Copy link
Member

tianon commented Dec 16, 2015 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

2 participants