On Fri, Dec 7, 2012 at 03:32:51PM +0100, Andres Freund wrote:
> > Well, it is a CONCURRENT index creation, so locking would be minimal.
>
> I wouldn't call a ShareUpdateExclusiveLock minimal...
>
> > Do we want pg_upgrade to be groveling over the lock view to look for
> > locks? I don't think so.
>
> ISTM that anybody who does DDL during or after pg_upgrade --check
> deserves any pain.
>
> So throwing an error in both seems perfectly fine for me.
Well, most of the current checks relate to checks for created objects.
To fail for in-process concurrent index creation is to fail for an
intermediate state --- index creation in process, but might complete
before we do the actual upgrade. Or it might not be an intermediate
state.
I am just saying that this makes the --check report more likely to false
failures than currently configured.
-- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB
http://enterprisedb.com
+ It's impossible for everything to be true. +