Re: Idea for improving buildfarm robustness - Mailing list pgsql-hackers

From Josh Berkus
Subject Re: Idea for improving buildfarm robustness
Date
Msg-id 560CB04B.3080704@agliodbs.com
Whole thread Raw
In response to Idea for improving buildfarm robustness  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
So, testing:

1. I tested running an AWS instance (Ubuntu 14.04) into 100% IOWAIT, and
the shutdown didn't kick in even when storage went full "d" state.  It's
possible that other kinds of remote storage failures would cause a
shutdown, but don't we want them to?

2. I tested deleting /pgdata/* several times (with pgbench running), and
Postgres shut down within 20 seconds each time.

3. I tested messing with the permissions on pg_control and global, and
Postgres threw other errors but continued running.

4. I mv'd the files and that didn't trigger a shutdown.

5. I did a fast swap:

rm -rf /pgdata/*
cp -p -r /pgdata2/* /pgdata/

... as expected, this did NOT cause postgres to shut down.

-- 
Josh Berkus
PostgreSQL Experts Inc.
http://pgexperts.com



pgsql-hackers by date:

Previous
From:
Date:
Subject: Re: [DOCS] max_worker_processes on the standby
Next
From: Kouhei Kaigai
Date:
Subject: Re: Parallel Seq Scan