Re: Filesystem fragmentation (Re: Fragmentation of WAL files)

From: Tom Lane
Subject: Re: Filesystem fragmentation (Re: Fragmentation of WAL files)
Date: ,
Msg-id: 26942.1177609772@sss.pgh.pa.us
(view: Whole thread, Raw)
In response to: Re: Filesystem fragmentation (Re: Fragmentation of WAL files)  (Gregory Stark)
List: pgsql-performance

Tree view

Fragmentation of WAL files  (Jim Nasby, )
 Re: Fragmentation of WAL files  (Bill Moran, )
  Re: Fragmentation of WAL files  (Heikki Linnakangas, )
   Filesystem fragmentation (Re: Fragmentation of WAL files)  (Bill Moran, )
    Re: Filesystem fragmentation (Re: Fragmentation of WAL files)  ("Craig A. James", )
     Re: Filesystem fragmentation (Re: Fragmentation of WAL files)  (Gregory Stark, )
      Re: Filesystem fragmentation (Re: Fragmentation of WAL files)  (Tom Lane, )
    Re: Filesystem fragmentation (Re: Fragmentation of WAL files)  (Florian Weimer, )
  Re: Fragmentation of WAL files  (Tom Lane, )
  Re: Fragmentation of WAL files  (Greg Smith, )

Gregory Stark <> writes:
> In the case of a performance-critical file like the WAL that's always read
> sequentially it may be to our advantage to defeat this technique and force it
> to be allocated sequentially. I'm not sure whether any filesystems provide any
> option to do so.

We more or less do that already by filling the entire file in one go
when it's created ...

            regards, tom lane


pgsql-performance by date:

From: "Alexander Staubo"
Date:
Subject: Re: Simple query, 10 million records...MySQL ten times faster
From: Mark Kirkwood
Date:
Subject: Re: postgres: 100% CPU utilization