| From: | "Imai, Yoshikazu" <imai(dot)yoshikazu(at)jp(dot)fujitsu(dot)com> |
|---|---|
| To: | 'David Rowley' <david(dot)rowley(at)2ndquadrant(dot)com> |
| Cc: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | RE: Small performance tweak to run-time partition pruning |
| Date: | 2018-11-08 00:34:00 |
| Message-ID: | 0F97FA9ABBDBE54F91744A9B37151A51210D11@g01jpexmbkw24 |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Mon, Oct 22, 2018 at 8:31 PM, David Rowley wrote:
> On 18 October 2018 at 16:13, Imai, Yoshikazu
> <imai(dot)yoshikazu(at)jp(dot)fujitsu(dot)com> wrote:
> > The patch improves the performance about 1.3% which is less than
> > David's result, but it seems still improves the performance.
>
> Thanks for doing these benchmarks.
>
> The speedup is small, but it becomes much more significant once other
> bottlenecks are removed. More partitions may show a larger increase, but
> more partitions also means that a larger range table array gets built
> during ExecInitRangeTable(), which is also slow.
Since I did enough tests and ISTM the patch is sophisticated,
I changed the state of this patch to "Ready for Committer".
--
Yoshikazu Imai
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Peter Geoghegan | 2018-11-08 01:46:52 | Re: Connections hang indefinitely while taking a gin index's LWLock buffer_content lock |
| Previous Message | Tom Lane | 2018-11-08 00:13:56 | Inadequate executor locking of indexes |