Re: [OT] linux 3.10 kernel will improve ipc,sysv semaphore scalability

From: Andrea Suisani <sickpig(at)opinioni(dot)net>
To: sthomas(at)optionshouse(dot)com
Cc: "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org>
Subject: Re: [OT] linux 3.10 kernel will improve ipc,sysv semaphore scalability
Date: 2013-05-16 08:47:07
Message-ID: 51949D0B.6030000@opinioni.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Hi all,

On 03/26/2013 07:59 PM, Shaun Thomas wrote:
> On 03/26/2013 08:04 AM, Andrea Suisani wrote:
>
>> TPS:
>> 100 users: 1257.21 (vanilla) 2805.06 (v3 patchset)
>> 400 users: 1437.57 (vanilla) 2664.67 (v3 patchset)
>> 800 users: 1236.89 (vanilla) 2750.73 (v3 patchset)
>
> Wow, I like the look of that. I wonder how this reacts with disabled autogrouping and increasing sched_migration_cost. If the completely fair scheduler has less locking contention with this patch-set, those tweaks may not even be necessary. I need to see if I can find a system to test on.
>

still no time to do proper testing but I just want to inform that the
patchset actually went in the 3.10 first release candidate, as we can see
from git log :)

git log v3.9..v3.10-rc1 --grep="ipc,sem: fine grained locking for semtimedop" -- ipc/sem.c

commit 6062a8dc0517bce23e3c2f7d2fea5e22411269a3
Author: Rik van Riel <riel(at)surriel(dot)com>
Date: Tue Apr 30 19:15:44 2013 -0700

ipc,sem: fine grained locking for semtimedop

...

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Shaun Thomas 2013-05-16 12:47:08 Re: Thinking About Correlated Columns (again)
Previous Message Craig James 2013-05-15 20:54:15 Re: Thinking About Correlated Columns (again)