Re: a heavy duty operation on an "unused" table kills my server

From: Greg Smith <greg(at)2ndquadrant(dot)com>
To: Eduardo Piombino <drakorg(at)gmail(dot)com>
Cc: Robert Haas <robertmhaas(at)gmail(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: a heavy duty operation on an "unused" table kills my server
Date: 2010-01-15 22:32:32
Message-ID: 4B50ED00.9090208@2ndquadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Eduardo Piombino wrote:
> Going to the disk properties (in windows), I just realized it does not
> have the Write Cache enabled, and it doesn't also allow me to set it
> up. I've read in google that the lack of ability to turn it on (that
> is, that the checkbox remains checked after you apply the changes),
> has to do with the lack of batter backup in the controller (which is
> default bundle option for embedded EA-200, which is our case).
>
> Regarding actual disk performance, I did some silly tests:
> Copied a 496 Mbytes file from a folder to another folder in C: and it
> took almost 90 secs.
> That would be 496MB/90 sec = 5.51MB/sec
>

I'd suggest http://www.hdtune.com/ as a better way to test transfer
speed here across the drive(s).

I think you'll find that your server continues to underperform
expectations until you get the battery installed that allows turning the
write cache on. A quick look at HP's literature suggests they believe
you only need the battery to enable the write-cache if you're using
RAID5. That's completely wrong for database use, where you will greatly
benefit from it regardless of underlying RAID setup. If you've got an
EA-200 but don't have a battery for it to unlock all the features,
you're unlikely to find a more cost effect way to improve your system
than to buy one.

--
Greg Smith 2ndQuadrant Baltimore, MD
PostgreSQL Training, Services and Support
greg(at)2ndQuadrant(dot)com www.2ndQuadrant.com

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Kenneth Marshall 2010-01-15 22:54:16 Bad plan choice nestloop vs. hashjoin
Previous Message Kevin Grittner 2010-01-15 21:55:08 Re: New server to improve performance on our large and busy DB - advice? (v2)