[rescue] Current collections...

[Patrick Finnegan]

On Monday 05 April 2004 13:05, Dave McGuire wrote:
> a couple of SS5/10/20 systems (one SS20 running in production)

Speaking of "in production".  I don't really have any problem with that 
sort of thing for production hardware, like a DNS, DHCP or other 
network service server.  In fact, I had a Sparc 5 running on top of 
240GB of raid storage a year or so ago.  However, it's not always a 
good idea...

The people that maintain "core network services" at Purdue until just 
recently had an SS20 in production that only got replaced because it 
had a memory problem, and then upon rebooting, its disk didn't spin 
back up.  The machine got replaced with an Ultra 5 (which is a 
reasonable machine, but not necessarily more stable than the SS20) and 
its service was back up the next morning.  It didn't (and still 
doesn't) have a hot spare.

Now, anyone dare to guess what service it was running?  

That's right, it was the print spooler for Purdue.  Not *a* print 
spooler, but *the* print spooler for most of campus.  Lots of people 
got pissed off when they couldn't print for a good 18 hours...

If they ever end up giving me a staff position here, I'll make sure I 
have hot spares for anything "critical" that I maintain. :)

Pat
-- 
Purdue University ITAP/RCS        ---  http://www.itap.purdue.edu/rcs/
The Computer Refuge               ---  http://computer-refuge.org