Re: Slow planning time when public schema included (12 vs. 9.4)

2020-03-24 Thread Anders Steinlein
On Sat, Mar 21, 2020 at 11:55 PM Tom Lane wrote: > Anders Steinlein writes: > > This they most definitely are not. 9.4 was running on an old box, Ubuntu > > 12.04, while 12 is on an up-to-date Ubuntu 18.04 LTS. AFAICS, 2.15 on the > > 9.4 box and 2.27 on the 12 box. > &

Re: Slow planning time when public schema included (12 vs. 9.4)

2020-03-21 Thread Anders Steinlein
On Sat, Mar 21, 2020 at 8:35 PM Tom Lane wrote: > Anders Steinlein writes: > > So I don't understand this big difference? Because it does seem like > citext > > is indeed the difference. > > It seems odd to me too. I'm not at all surprised that citext compari

Re: Slow planning time when public schema included (12 vs. 9.4)

2020-03-21 Thread Anders Steinlein
On Sat, Mar 21, 2020 at 3:26 PM Tom Lane wrote: > Anders Steinlein writes: > > We haven't noticed any issues with this before now, until we started > seeing > > really slow planning time on some relatively simple queries: > > ... > > The planning time is the

Re: Slow planning time when public schema included (12 vs. 9.4)

2020-03-21 Thread Anders Steinlein
On Sat, Mar 21, 2020 at 2:37 PM Andreas Kretschmer wrote: > > > Am 21.03.20 um 13:02 schrieb Anders Steinlein: > > default_statistics_target = 1000 > > not sure if this be the culprit here, but i think this is way too high. > Leave it at the normal value of 100 and rais

Slow planning time when public schema included (12 vs. 9.4)

2020-03-21 Thread Anders Steinlein
Hi, We have a multi-tenant app where each tenant has their own schemas, and inherits tables from tables defined in the public schema. Other shared data such as data types are also stored in the public schema. While running this app, every transaction is started with setting the search_path to , pu