Re: s_lock() seems too aggressive for machines with many sockets

From: Andres Freund <andres(at)anarazel(dot)de>
To: Nils Goroll <slink(at)schokola(dot)de>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Jan Wieck <jan(at)wi3ck(dot)info>, PostgreSQL Development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: s_lock() seems too aggressive for machines with many sockets
Date: 2015-06-10 15:01:25
Message-ID: 20150610150125.GC10551@awork2.anarazel.de
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

On 2015-06-10 16:55:31 +0200, Nils Goroll wrote:
> But still I am convinced that on today's massively parallel NUMAs, spinlocks are
> plain wrong:

Sure. But a large number of installations are not using massive NUMA
systems, so we can't focus on optimizing for NUMA.

We definitely have quite some catchup to do there. Unfortunately most of
the problems are only reproducible on 4, 8 socket machines, and it's
hard to get hand on those for prolonged amounts of time.

> - Even if critical sections are kept minimal, they can still become hot spots

That's why we started to remove several of them...

> - The fact that well behaved mutexes have a higher initial cost could even
> motivate good use of them rather than optimize misuse.

Well. There's many locks in a RDBMS that can't realistically be
avoided. So optimizing for no and moderate contention isn't something
you can simply forgo.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jan Wieck 2015-06-10 15:03:31 Re: s_lock() seems too aggressive for machines with many sockets
Previous Message Gurjeet Singh 2015-06-10 15:00:28 Re: replication slot restart_lsn initialization