Re: track_planning causing performance regression

On 2020/07/04 12:22, Pavel Stehule wrote:
>
>
> pá 3. 7. 2020 v 13:02 odesílatel Fujii Masao <masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>> napsal:
>
>
>
>     On 2020/07/03 16:02, Pavel Stehule wrote:
>      >
>      >
>      > pá 3. 7. 2020 v 8:57 odesílatel Fujii Masao <masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>>> napsal:
>      >
>      >
>      >
>      >     On 2020/07/03 13:05, Pavel Stehule wrote:
>      >      > Hi
>      >      >
>      >      > pá 3. 7. 2020 v 4:39 odesílatel Fujii Masao <masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>>>> napsal:
>      >      >
>      >      >
>      >      >
>      >      >     On 2020/07/01 7:37, Peter Geoghegan wrote:
>      >      >      > On Tue, Jun 30, 2020 at 6:40 AM Fujii Masao <masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com> <mailto:masao.fujii@oss.nttdata.com <mailto:masao.fujii@oss.nttdata.com>>>> wrote:
>      >      >      >> Ants and Andres suggested to replace the spinlock used in pgss_store() with
>      >      >      >> LWLock. I agreed with them and posted the POC patch doing that. But I think
>      >      >      >> the patch is an item for v14. The patch may address the reported performance
>      >      >      >> issue, but may cause other performance issues in other workloads. We would
>      >      >      >> need to measure how the patch affects the performance in various workloads.
>      >      >      >> It seems too late to do that at this stage of v13. Thought?
>      >      >      >
>      >      >      > I agree that it's too late for v13.
>      >      >
>      >      >     Thanks for the comment!
>      >      >
>      >      >     So I pushed the patch and changed default of track_planning to off.
>      >      >
>      >      >
>      >      > Maybe there can be documented so enabling this option can have a negative impact on performance.
>      >
>      >     Yes. What about adding either of the followings into the doc?
>      >
>      >           Enabling this parameter may incur a noticeable performance penalty.
>      >
>      >     or
>      >
>      >           Enabling this parameter may incur a noticeable performance penalty,
>      >           especially when a fewer kinds of queries are executed on many
>      >           concurrent connections.
>      >
>      >
>      > This second variant looks perfect for this case.
>
>     Ok, so patch attached.
>
>
> +1

Thanks for the review! Pushed.

Regards,

--
Fujii Masao
Advanced Computing Technology Center
Research and Development Headquarters
NTT DATA CORPORATION