Re: databases and RAID ... - Mailing list pgsql-admin

From Ragnar Kjørstad
Subject Re: databases and RAID ...
Date
Msg-id 20020525220119.A32683@vestdata.no
Whole thread Raw
In response to Re: databases and RAID ...  ("Fred Moyer" <fred@digicamp.com>)
List pgsql-admin
On Sat, May 25, 2002 at 12:45:12PM -0700, Fred Moyer wrote:
> Performance (fastest->slowest)
>     hardware raid -> software raid
>     raid 0 -> 10 -> 1 -> 5
> Redundancy (most -> least)
>     hardware raid -> software raid
>     10, 1 -> 5 -> 0

It's really not possible to compare RAID-levels independent from what
the system is beeing used for. E.g. lots of seeks vs continous access,
read-intensive vs write-intensive, how many simultanious accesses and
so on.

E.g. RAID 1 / 10 can easily be as fast, or faster than RAID 0 for read
intensive work.

RAID 5 has a very high penality when doing small writes, but the effect
can be reduced by good RAID-controllers with lots of battery-backed
cached.


For a typical database-application I would agree with your statement
except that RAID 1 is probably faster than RAID 10.


--
Ragnar Kjorstad
Big Storage

pgsql-admin by date:

Previous
From: Andy Ruhl
Date:
Subject: Re: databases and RAID ...
Next
From: Jyry Kuukkanen
Date:
Subject: Re: How to change datatype of a field