| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Andres Freund <andres(at)anarazel(dot)de> |
| Cc: | Peter Geoghegan <pg(at)heroku(dot)com>, Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>, Pg Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: rethinking dense_alloc (HashJoin) as a memory context |
| Date: | 2016-07-13 17:56:00 |
| Message-ID: | 22626.1468432560@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Andres Freund <andres(at)anarazel(dot)de> writes:
> On 2016-07-13 13:37:55 -0400, Tom Lane wrote:
>> We already know that
>> that code has performance issues, cf bug #14231, so I suspect there's
>> a redesign in its future anyway.
> Note that it's not the slab allocators that is having problems, it's
> aset.c, iterating through all allocated blocks.
Well, my point is that that code is using allocation patterns that aset.c
isn't very well suited for. The slab allocator logic attempts to paper
over one part of that impedance mismatch, but we're seeing there's more.
I haven't looked at it closely enough to have a solution proposal.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Corey Huinker | 2016-07-13 18:06:20 | Re: \timing interval |
| Previous Message | Andres Freund | 2016-07-13 17:48:38 | Re: Re: [COMMITTERS] pgsql: Avoid extra locks in GetSnapshotData if old_snapshot_threshold < |