Re: pg_upgrade and rsync

On Tue, Mar 3, 2015 at 08:38:50AM -0500, Bruce Momjian wrote:
> > < 2015-02-24 11:47:22.861 MSK >WARNING: WAL was generated with wal_level=
> > minimal, data may be missing
> > < 2015-02-24 11:47:22.861 MSK >HINT: This happens if you temporarily set
> > wal_level=minimal without taking a new base backup.
> > < 2015-02-24 11:47:22.861 MSK >FATAL: hot standby is not possible because
> > wal_level was not set to "hot_standby" or higher on the master server
> > < 2015-02-24 11:47:22.861 MSK >HINT: Either set wal_level to "hot_standby" on
> > the master, or turn off hot_standby here.
> > < 2015-02-24 11:47:22.862 MSK >LOG: startup process (PID 28093) exited with
> > exit code 1
> > < 2015-02-24 11:47:22.862 MSK >LOG: aborting startup due to startup process
> > failure
>
> OK, hmmm. Thanks for testing. It feels like you didn't have your new
> master set up for streaming replication when you ran pg_upgrade. Is
> that correct? Should I specify that specifically in the instructions?

Actually, I think you are on to something that needs to be documented.
Because the old and new clusters might be using the same port number,
you can't configure the new master to use streaming replication because
you can't be shipping those logs to the old standby. Yikes. OK, I
think we need to document that you need to set wal_level=hot_standby on
the new master, but not set up streaming. Once you are done the
upgrade, you should configure streaming.

If this fixes the problem, I will generate an updated documentation
patch. Please let me know.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ Everyone has their own god. +