aic7896 SCB timeout - is this a sign of impending doom?

[Doug White]

On Thu, 23 Oct 2003, Scott Mitchell wrote:

> The drives are housed in a hot-swap cage in an Intel server case, so cabling
> or termination problems would be quite serious... there's only one cable and
> that's hardwired in.  The drives are ~3 years old so it would not surprise
> me if one was on the way out.  Might be time to investigate the SMART
> monitoring tools that were mentioned on here a week or so ago.

Yeah, it wouldn't be the first time SCSI backplanes have gone bad.  You
have this in an Astor or Columbus chassis?

> Temperature shouldn't be a problem given the number of fans in the case,
> but I'll check that they're all still running OK.  This particular box is
> at the bottom of a rack in a room with a ridiculous oversupply of underfloor
> aircon - overheating has never been a problem here :-)

You never know.  Older half-height drives (like the IBM DMVS series) get
REALLY REALLY HOT, and if a fan has gone out it could cause serious
cooling issues.

> Agreed, they're excellent machines.  We use t pair of them as file / cvs /
> DNS / NIS / www / etc. servers, which they're more than adequate for.

Unfortunately they run a really old version of the IPMI spec, otherwise I
have some scripts that can inquiry for temperature data.  Maybe sometime
I'll get bored and backport the stuff to IPMI 0.9.

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org