Re: Linux I/O tuning: CFQ vs. deadline

From: Albe Laurenz
Subject: Re: Linux I/O tuning: CFQ vs. deadline
Date: ,
Msg-id: D960CB61B694CF459DCFB4B0128514C203938157@exadv11.host.magwien.gv.at
(view: Whole thread, Raw)
In response to: Linux I/O tuning: CFQ vs. deadline  (Greg Smith)
Responses: Re: Linux I/O tuning: CFQ vs. deadline  ("Kevin Grittner")
List: pgsql-performance

Tree view

Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
 Re: Linux I/O tuning: CFQ vs. deadline  ("Albe Laurenz", )
  Re: Linux I/O tuning: CFQ vs. deadline  ("Kevin Grittner", )
   Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
    Re: Linux I/O tuning: CFQ vs. deadline  (Josh Berkus, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Scott Marlowe, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Mark Wong, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Scott Carey, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Jeff Davis, )
 Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
  Re: Linux I/O tuning: CFQ vs. deadline  (, )
   Re: Linux I/O tuning: CFQ vs. deadline  (Jeff, )
    Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Scott Marlowe, )
      Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
       Re: Linux I/O tuning: CFQ vs. deadline  (, )
        Re: Linux I/O tuning: CFQ vs. deadline  (Greg Smith, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Jeff, )
     Re: Linux I/O tuning: CFQ vs. deadline  (Scott Carey, )

Greg Smith wrote:
> Recently I've made a number of unsubstantiated claims that the deadline
> scheduler on Linux does bad things compared to CFQ when running
> real-world mixed I/O database tests.  Unfortunately every time I do one
> of these I end up unable to release the results due to client
> confidentiality issues.  However, I do keep an eye out for people who
> run into the same issues in public benchmarks, and I just found one:
> http://insights.oetiker.ch/linux/fsopbench/

That is interesting; particularly since I have made one quite different
experience in which deadline outperformed CFQ by a factor of approximately 4.

So I tried to look for differences, and I found two possible places:
- My test case was read-only, our production system is read-mostly.
- We did not have a RAID array, but a SAN box (with RAID inside).

The "noop" scheduler performed about as well as "deadline".
I wonder if the two differences above could explain the different
result.

Yours,
Laurenz Albe


pgsql-performance by date:

From: Andres Freund
Date:
Subject: Re: [HACKERS] Re: Faster CREATE DATABASE by delaying fsync (was 8.4.1 ubuntu karmic slow createdb)
From: Mark Wong
Date:
Subject: Re: Linux I/O tuning: CFQ vs. deadline