Horton

Re: Errors creating partitioned tables from existing using (LIKE <table>) after renaming table constraints

Amit Langote <langote_amit_f8@lab.ntt.co.jp>

View thread View on postgr.es
From: Amit Langote <Langote_Amit_f8@lab.ntt.co.jp>
To: Michael Paquier <michael@paquier.xyz>
Cc: Stuart <sfbarbee@gmail.com>, pgsql-bugs@postgresql.org
Date: 2018-12-17T02:20:14Z
Lists: pgsql-bugs 
On 2018/12/17 11:06, Michael Paquier wrote:
> On Mon, Dec 17, 2018 at 10:51:10AM +0900, Michael Paquier wrote:
>> Done just a few moments ago.  I have also added a test with a primary
>> key rename which does not matter as long as LIKE INCLUDE INDEXES is
>> involved but that felt safer in the long-run.
> 
> Buildfarm member prion is complaining on that:
> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=prion&dt=2018-12-17%2001%3A46%3A36
> This uses -DRELCACHE_FORCE_RELEASE -DCATCACHE_FORCE_RELEASE.
> 
> It seems that the call to relation_close should happen after
> CacheInvalidateRelcache.  Damn..

What a blunder!  Sorry. :(

Thanks,
Amit

Commits

  1. Make constraint rename issue relcache invalidation on target relation