Re: Perf Benchmarking and regression.

Поиск
Список
Период
Сортировка
От Mithun Cy
Тема Re: Perf Benchmarking and regression.
Дата
Msg-id CAD__OuiObzNVTt_hO__P5AEnU4iNqcFWgArXR4TbLKe-UXyukQ@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Perf Benchmarking and regression.  (Andres Freund <andres@anarazel.de>)
Ответы Re: Perf Benchmarking and regression.  (Andres Freund <andres@anarazel.de>)
Список pgsql-hackers
Дерево обсуждения

On Fri, May 6, 2016 at 8:35 PM, Andres Freund <andres@anarazel.de> wrote:
> Also, do you see read-only workloads to be affected too?
Thanks, I have not tested with above specific commitid which reported performance issue but
At HEAD commit 72a98a639574d2e25ed94652848555900c81a799
Author: Andres Freund <andres@anarazel.de>
Date:   Tue Apr 26 20:32:51 2016 -0700

READ-Only (prepared) tests (both when data fits to shared buffers or it exceeds shared buffer=8GB) performance of master has improved over 9.5

SessionsPostgreSQL-9.5 scale 300PostgreSQL-9.6 scale 300%diff
15287.5615945213.723197-1.396454598
884265.38908384871.3056890.719057507
16148330.4155158661.1283156.9646624936
24207062.803697219958.129746.2277366155
32265145.089888290190.5014439.4459269699
40311688.752973340000.5517729.0833559212
48327169.9673372408.07303313.8270960829
56274426.530496390629.2494842.3438356248
64261777.692042384613.966646.9238893505
72210747.55937376390.16202278.5976374517
80220192.818648398128.77932980.8091570713
88185176.91888423906.711882128.9198429512
96161579.719039421541.656474160.8877271115
104146935.568434450672.740567206.7145316618
112136605.466232432047.309248216.2738074582
120127687.175016455458.086889256.6983816753
128120413.936453428127.879242255.5467845776
 
SessionsPostgreSQL-9.5 scale 1000PostgreSQL-9.6 scale 1000%diff
15103.8122025155.4348081.01145191
847741.904153117.80509611.2603405694
1689722.5703186965.10079-3.0733287182
24130914.537373153849.63424517.5191367836
32197125.725706212454.4742647.7761279017
40248489.551052270304.0937678.7788571482
48291884.652232317257.8367468.6928806705
56304526.216047359676.78547618.1102862489
64301440.463174388324.71018528.8230206709
72194239.941979393676.628802102.6754254511
80144879.527847383365.678053164.6099719885
88122894.325326372905.436117203.4358463076
96109836.31148362208.867756229.7715144249
104103791.981583352330.402278239.4582094921
112105189.206682345722.499429228.6672752217
120108095.811432342597.969088216.939171416
128113242.59492333821.98763194.7848270925

Even for READ-WRITE when data fits into shared buffer (scale_factor=300 and shared_buffers=8GB) performance has improved.
Only case is when data exceeds shared_buffer(scale_factor=1000 and shared_buffers=8GB) I see some regression.

I will try to run the tests as you have suggested and will report the same.
 

Thanks and Regards
Mithun C Y

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Vladimir Gordiychuk
Дата:
Сообщение: Stopping logical replication protocol
Следующее
От: Robert Haas
Дата:
Сообщение: Re: SET ROLE and reserved roles