>
> Presumably I could partition proposal on has_been_anonymised, however
> the row counts seem low enough that it feels a bit like overkill? We
> also need referential integrity so I'll need to wait until that's in
> (I think it's coming in PG12?)
>
> If I decrease the number of
Aha!
That's a great hint, we had that set down to an obscenely low value
due to our max_connections setting being quite high. I've tweaked it
back up to 4MB for now and it's definitely had a marked improvement!
Many Thanks,
Rob
On 09/08/2019, Jeff Janes wrote:
> On Fri, Aug 9, 2019 at 4:42 AM
On Fri, Aug 9, 2019 at 4:42 AM Rob Emery wrote:
>
> It
> seems to me like the Bitmap Heap Scan on proposal is the issue because
> the recheck is throwing away enormous amounts of data.
Have you tried increasing work_mem? The probable reason for the recheck is
that your bitmap overflows the
Hi Guys,
I’m at a bit of a loss where I can go with the following 2 queries
that are over the same data structure (DDL attached) under postgresql
PostgreSQL 9.5.16 on x86_64-pc-linux-gnu (Debian 9.5.16-1.pgdg90+1),
compiled by gcc (Debian 6.3.0-18+deb9u1) 6.3.0 20170516, 64-bit and
could do with