Re: raid10 write performance

From: Karl Denninger
Subject: Re: raid10 write performance
Date: ,
Msg-id: 4C20C8E1.7060306@denninger.net
(view: Whole thread, Raw)
In response to: Re: raid10 write performance  (Justin Graf)
Responses: Re: raid10 write performance  (Dave Crooke)
Re: raid10 write performance  (Scott Carey)
List: pgsql-performance

Tree view

raid10 write performance  (Grzegorz Jaśkiewicz, )
 Re: raid10 write performance  (Justin Graf, )
  Re: raid10 write performance  (Karl Denninger, )
   Re: raid10 write performance  (Dave Crooke, )
   Re: raid10 write performance  (Scott Carey, )
 Re: raid10 write performance  (Greg Smith, )
  Re: raid10 write performance  (Ivan Voras, )
   Re: raid10 write performance  (Florian Weimer, )
    Re: raid10 write performance  (Ivan Voras, )
     Re: raid10 write performance  (Matthew Wakeling, )
     Re: raid10 write performance  (Scott Marlowe, )
   Re: raid10 write performance  (Scott Marlowe, )

Justin Graf wrote:
On 6/22/2010 4:31 AM, Grzegorz Jaśkiewicz wrote: 
Would moving WAL dir to separate disk help potentially ?     
Yes it can have a big impact.
WAL on a separate spindle will make a HUGE difference in performance.  TPS rates frequently double OR BETTER with WAL on a dedicated spindle.

Strongly recommended.

Be aware that you must pay CLOSE ATTENTION to your backup strategy if WAL is on a different physical disk.  Snapshotting the data disk where WAL is on a separate spindle and backing it up **WILL NOT WORK** and **WILL** result in an non-restoreable backup.

The manual discusses this but it's easy to miss.... don't or you'll get a NASTY surprise if something goes wrong.....

-- Karl
Attachment

pgsql-performance by date:

From: Scott Carey
Date:
Subject: ALTER Table and CLUSTER does adding a new column rewrite clustered? (8.4.3)
From: Anj Adu
Date:
Subject: slow index lookup