| From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
|---|---|
| To: | Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com> |
| Cc: | Peter Geoghegan <pg(at)bowt(dot)ie>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Thomas Munro <thomas(dot)munro(at)enterprisedb(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Boom filters for hash joins (was: A design for amcheck heapam verification) |
| Date: | 2017-09-20 02:01:54 |
| Message-ID: | CA+TgmoZGVH8CWz2PT8ndVmn8-6no3Pa=W7ujny-j=2v=rV49LQ@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Tue, Sep 19, 2017 at 4:25 PM, Tomas Vondra
<tomas(dot)vondra(at)2ndquadrant(dot)com> wrote:
> I haven't thought about it from that point of view. Can you elaborate
> why that would be the case? Sorry if this was explained earlier in this
> thread (I don't see it in the history, though).
>
> I can't quite remember why I haven't pursued the patch in 2015, but it
> was probably clear it wouldn't get in in the last CF, and I never got
> back to it.
IIRC, it was a clear loser performance-wise in the case where the
Bloom filter didn't end up helping, and we didn't have a way to avoid
doing it when it didn't help. That may or may not be why you didn't
pursue it, but I'm fairly sure it was my motivation for being
unexcited about the whole idea. I think if we can solve that problem
somehow, we have a winner.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Masahiko Sawada | 2017-09-20 02:07:31 | Re: Commits don't block for synchronous replication |
| Previous Message | Peter Geoghegan | 2017-09-20 02:01:47 | Re: CREATE COLLATION does not sanitize ICU's BCP 47 language tags. Should it? |