Re: postgres files in use not staying in linux file cache

From: Tim Kane <tim(dot)kane(at)gmail(dot)com>
To: Shaun Thomas <sthomas(at)optionshouse(dot)com>, Brio <brianoraas(at)gmail(dot)com>, <pgsql-performance(at)postgresql(dot)org>
Subject: Re: postgres files in use not staying in linux file cache
Date: 2014-06-13 07:19:19
Message-ID: CFC06009.865FB%tim.kane@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance


>
> From: Shaun Thomas <sthomas(at)optionshouse(dot)com>
> Date: Tuesday, 10 June 2014 22:07
> So here's the thing. The Linux page reclamation code is *extremely
> broken* in everything before 3.11. Take a look at this, then realize
> that this is *only one patch* from several that target the memory
> manager weightings:
>
> <snipped>
>
> Since you're using Ubuntu 12.04, I strongly suggest upgrading your core
> to 12.04.4 and apply the linux-generic-lts-saucy pseudo-package to at
> least get onto the 3.11 instead. The 3.2 kernel is pants-on-head
> retarded; we've had a lot more luck with 3.8 and above.

>
Without trying to hijack this thread, is there any documentation around
recommended or known-good kernels for Postgres/DB type workloads?
I can see there has been a lot of discussion around recent kernel
developments being detrimental (or potentially detrimental) -
http://www.postgresql.org/message-id/20140115141408.GJ4963@suse.de - , but
there doesn’t seem to be a definitive resource to know which kernels to
avoid.

I ask because I’ve just read your statement above about 3.2 being
pants-on-head, and having had more luck with 3.8 and above – despite most
installations being on much older (2.6.19) kernels (as per the thread).
I’d be interested to see how much benefit we’d get from moving off 3.2, but
I’d like to be aware of the risks of more recent kernel version’s also.

Cheers,

TIm

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Shaun Thomas 2014-06-13 16:11:41 Re: postgres files in use not staying in linux file cache
Previous Message Matt Silverlock 2014-06-12 22:39:43 Re: OFFSET/LIMIT - Disparate Performance w/ Go application