From: | David Conlin <dc345(at)cantab(dot)net> |
---|---|
To: | Adrien NAYRAT <adrien(dot)nayrat(at)anayrat(dot)info>, pgsql-bugs(at)lists(dot)postgresql(dot)org |
Subject: | Re: BUG #15614: Query plan: buffer stats from workers in child operations discarded. |
Date: | 2019-01-31 14:29:40 |
Message-ID: | 1f5ac84b-1899-b908-746d-655a0abe3d87@cantab.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-bugs |
It is, thanks a lot.
Sorry, I should have checked - I thought the minor versions were getting
bugfixes, but I was just confused with the pre v10 way of numbering I think
Thanks
Dave
On 31/01/2019 10:28, Adrien NAYRAT wrote:
> On 1/31/19 10:33 AM, PG Bug reporting form wrote:
>> The following bug has been logged on the website:
>>
>> Bug reference: 15614
>> Logged by: Dave Conlin
>> Email address: dc345(at)cantab(dot)net
>> PostgreSQL version: 10.4
>> Operating system: Windows 10
>> Description:
>>
>> Hi folks -
>>
>> In the output of an `EXPLAIN` query, normally child operations are
>> included
>> in the totals for the buffers values, eg "Shared Hit Blocks". It
>> seems that
>> when a parent operation is single-threaded, but a child operation starts
>> workers, then only the buffers stats from the main thread in the child
>> operation are included in the totals for the parent operation.
>>
>
>
> Hello,
>
> This should have been fixed in 10.5 :
>
> https://www.postgresql.org/docs/current/release-10-5.html
> > Fix EXPLAIN's accounting for resource usage, particularly buffer
> accesses, in parallel workers (Amit Kapila, Robert Haas)
>
> Could you apply the minor update and test again?
>
> Regards,
>
From | Date | Subject | |
---|---|---|---|
Next Message | Bruce Momjian | 2019-01-31 14:35:58 | Re: BUG #15609: synchronous_commit=off insert performance regression with secondary indexes |
Previous Message | Kyotaro HORIGUCHI | 2019-01-31 12:26:48 | Re: BUG #15589: Due to missing wal, restore ends prematurely and opens database for read/write |