Re: Move PinBuffer and UnpinBuffer to atomics
От | Dilip Kumar |
---|---|
Тема | Re: Move PinBuffer and UnpinBuffer to atomics |
Дата | |
Msg-id | CAFiTN-tF00jA5yVrpq9ZipbenvszpWWh1fRdp8ZUZX1QVGvEJQ@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Move PinBuffer and UnpinBuffer to atomics (Andres Freund <andres@anarazel.de>) |
Ответы |
Re: Move PinBuffer and UnpinBuffer to atomics
|
Список | pgsql-hackers |
On Fri, Apr 1, 2016 at 2:09 PM, Andres Freund <andres@anarazel.de> wrote:
One interesting thing to do would be to use -P1 during the test and see
how much the performance varies over time.
I have run with -P option, I ran for 1200 second and set -P as 30 second, and what I observed is that when its low its low throughout the run and when its high, Its high for complete run.
Non Default Parameter:
--------------------------------
shared_buffer 8GB
Max Connections 150
./pgbench -c $threads -j $threads -T 1200 -M prepared -S -P 30 postgres
Test results are attached in result.tar file.
File details:
------------------
1. head_run1.txt --> 20 mins run on head reading 1
2. head_run2.txt --> 20 mins run on head reading 2
3. head_patch_run1.txt --> 20 mins run on head + patch reading 1
4. head_patch_run2.txt --> 20 mins run on head + patch reading 2
5. head_pinunpin.txt --> 20 mins run on head + pinunpin-cas-8.patch
6. head_pinunpin_patch.txt --> 20 mins run on head + pinunpin-cas-8.patch + patch reading
--
Вложения
В списке pgsql-hackers по дате отправления: