[plug] Server stress test, RAID health monitoring

[Gavin Chester]

On Sat, 2005-06-11 at 17:53 +0800, Matthew Lambie wrote:
> I have received three new servers for work and I'm installing the OS
> over the weekend. We're using Ubuntu on all the machines, and they all
> have SATA RAID configurations. I love the smell of new hardware :)
> 
> 1 - Can anyone recommend a good way/application to really hammer the
> processor, memory and disk? In the past I've compiled kernel after
> kernel in a loop but there must be a suite of tools people use to "burn
> in" new machines. If the disk is going to die, it will in the first few
> days, so I'd like to thump it now and replace it before it goes into the
> racks next week.

I remember seeing an app that was designed to do just that - but can't
remember its name, sorry.  Something like "stresstest", funny enough.
Perhaps google will find it, I stumbled on it at Tucows or some such
s/ware repository. Sorry not more helpful.

Gavin

> 
> 2 - What do people use to monitor the health of RAID arrays? I want to
> know if the array is dying so we can get a new disk in, obviously. I'm
> expecting that tools exist which will fire off emails when problems are
> found, but if not I can happily script something triggered from cron.
> We're using Adaptec 2410SA SATA RAID cards, if that has any bearing on
> things.
>