Re: Update maintenance_work_mem/autovacuum_work_mem to reflect the 1GB limitation with VACUUM

On Wed, 7 Jul 2021 at 23:44, David Rowley <dgrowleyml(at)gmail(dot)com> wrote:
>
> On Sun, 4 Jul 2021 at 22:38, David Rowley <dgrowleyml(at)gmail(dot)com> wrote:
> > I could do with a 2nd opinion about if we should just adjust the
> > maximum value for the autovacuum_work_mem GUC to 1GB in master.
> >
> > I'm also not sure if since we'd not backpatch the GUC max value
> > adjustment if we need to document the upper limit in the manual.
>
> I was just looking at this again and I see that GIN indexes are able
> to use more than 1GB of memory during VACUUM. That discovery makes me
> think having the docs say that vacuum cannot use more than 1GB of
> memory is at best misleading and more likely just incorrect.

The attached patch aims to put right where I went wrong with the
documentation about vacuum/autovacuum only using maintainance_work_mem
memory for dead tuple collection.

I plan to push this and backpatch to 9.6 shortly unless there are any
better ideas.

What's in there right now is wrong and I want that fixed before the
cut-off for the next set of bug fix releases.

David