Re[2]: [BUGS] Re[2]: [BUGS] BUG #13869: Right Join query that never ends

From: Master ZX <zx-master(at)bigmir(dot)net>
To: Peter Geoghegan <pg(at)heroku(dot)com>
Cc: David G(dot) Johnston <david(dot)g(dot)johnston(at)gmail(dot)com>, pgsql-bugs(at)postgresql(dot)org <pgsql-bugs(at)postgresql(dot)org>
Subject: Re[2]: [BUGS] Re[2]: [BUGS] BUG #13869: Right Join query that never ends
Date: 2016-01-26 11:49:42
Message-ID: E1aO27z-0001JM-0L@bst04.sputnikmedia.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

19.01.2016 02:47, Peter Geoghegan <pg(at)heroku(dot)com>
>On Sat, Jan 16, 2016 at 4:51 AM, Master ZX <zx-master(at)bigmir(dot)net> wrote:
> > I can do more to investigate this behavior only if postgresql will consider my claim as bug that has to be fixed in first order.
> > Ideally I should give my DB schema and dump or VM wich reproduces this bug, but I can not give away our internal data. All other options I can not describe at public mailing list.
>
> Why don't you use a tool like Linux's perf to determine where time is spent?:
>
> https://wiki.postgresql.org/wiki/Profiling_with_perf
>
> A simple "perf top" may be all that it takes to give us a better idea
> of where time is spent.
>
> --
> Peter Geoghegan

This is forwarded message from 21 of Jan (looks like previous message lost)

I'm not really experienced in bug reporting and perf usage.
I've recorded some data via perf record for about 40 sec. and this is what I was able to get via perf report -g --children

- 8.98% 8.93% postgres postgres [.] slot_getattr ▒
slot_getattr ▒
- 5.79% 5.75% postgres postgres [.] ExecMergeJoin ▒
ExecMergeJoin ▒
- 3.73% 3.70% postgres postgres [.] ExecQual ▒
ExecQual ▒
- 3.68% 3.65% postgres postgres [.] ExecStoreMinimalTuple ▒
ExecStoreMinimalTuple ▒
- 2.85% 0.00% postgres [unknown] [.] 0x0000000000000001 ▒
0x1 ▒
- 2.80% 2.78% postgres postgres [.] tuplesort_gettupleslot ▒
tuplesort_gettupleslot ▒
- 2.78% 0.00% postgres postgres [.] 0xffffffffffdb3602 ▒
0x1b3602 ▒
- 2.76% 2.76% postgres postgres [.] 0x00000000001b3602 ▒
0x1b3602 ▒
- 2.64% 2.62% postgres postgres [.] MemoryContextReset ▒
MemoryContextReset ▒
- 2.30% 2.28% postgres postgres [.] ExecProcNode ▒
ExecProcNode ▒
- 2.26% 2.25% postgres postgres [.] pgstat_end_function_usage ▒
pgstat_end_function_usage ▒
- 2.24% 0.00% postgres postgres [.] 0xffffffffffdb35f0 ▒
0x1b35f0 ▒
- 2.23% 2.23% postgres postgres [.] 0x00000000001b35f0 ▒
0x1b35f0 ▒
- 2.13% 2.12% postgres postgres [.] ExecSort ▒
ExecSort ▒
- 1.87% 1.86% postgres postgres [.] check_stack_depth ▒
+ check_stack_depth ▒
- 1.47% 1.46% postgres postgres [.] pgstat_init_function_usage ▒
pgstat_init_function_usage

I have about 100Mb of perf.data and as far as I can understand perf report will be most informative on machine where it was collected.
So in a case if you need some other data from perf.dump just let me know.

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message onderkalaci 2016-01-26 12:34:20 BUG #13891: Deparsed arbiter WHERE clauses cannot be parsed by Postgres
Previous Message Dmitriy Sarafannikov 2016-01-26 09:53:50 Re: [BUGS] Postgres backend segfault