Matthew Wakeling wrote:
> In fact, it is recommended that two different types of drives are used.
> That way, if there's a mass failure of a whole batch of drives from one
> particular vendor, you don't lose all your data.
Don't think this is just paranoia, either. I've had it happen to me
SEVERAL times - either a second drive fails before I can replace the
first, or the second drive in a pair fails during rebuild onto the
replacement. I use regular RAID scrubbing, so this isn't just a case of
undetected media errors.
Nothing beats good backups.
In response to
pgsql-performance by date
|Next:||From: Craig Ringer||Date: 2009-02-14 08:49:05|
|Subject: Re: I/O increase after upgrading to 8.3.5|
|Previous:||From: Chris Browne||Date: 2009-02-13 17:37:21|
|Subject: Re: scheduling autovacuum at lean hours only.|