From: | Merlin Moncure <mmoncure(at)gmail(dot)com> |
---|---|
To: | Jeff Janes <jeff(dot)janes(at)gmail(dot)com> |
Cc: | sthomas(at)optionshouse(dot)com, Vlad <marchenko(at)gmail(dot)com>, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>, pgsql-general(at)postgresql(dot)org |
Subject: | Re: High SYS CPU - need advise |
Date: | 2012-11-20 22:52:15 |
Message-ID: | CAHyXU0yBshNe1OzVEDVNsMS93fsFQzAautirQ-wE_omtB0RH5g@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Tue, Nov 20, 2012 at 4:08 PM, Jeff Janes <jeff(dot)janes(at)gmail(dot)com> wrote:
>> It strikes me as cavalier to be resetting
>> trycounter while sitting under the #1 known contention point for read
>> only workloads.
>
> The only use for the trycounter is to know when to ERROR out with "no
> unpinned buffers available", so not resetting that seems entirely
> wrong.
>
> I would contest "the #1 known contention point" claim. We know that
> the freelist lock is a point of contention under certain conditions,
> but we (or at least I) also know that it is the mere acquisition of
> this lock, and not the work done while it is held, that is important.
Well, I'm speculating (without evidence) if that is in fact might not
be always true. It certainly is hard to demonstrate with synthetic
testing though. I guess the first step would be to whip up a
diagnostic patch to confirm malfeasance in suspicious cases.
merlin
From | Date | Subject | |
---|---|---|---|
Next Message | Gavin Flower | 2012-11-20 23:06:48 | Re: High SYS CPU - need advise |
Previous Message | Shaun Thomas | 2012-11-20 22:41:21 | Re: High SYS CPU - need advise |