What happened during upgrade?
Well, I have spent the last 2 days recovering from an upgrade to our
mainframe. We had an H30 with a 60mip processor on the CP side and a 120mip
processor on the IFL side. We had an install of Woody on the IFL side that
was running LPRng and Samba for our print server. Over the weekend we
upgraded our machine to an H50. We upgraded the 60mip processor on the CP
side to 120mip, doubled our disk space, and upgraded the microcode. The
added disk space is not yet formatted.
So, Monday morning came around and our mainframe seemed to be running great
with the upgrade. That was except for the Debian Print Server. For some
reason, the print server was processing items in the queue about 1 per 30
minutes. By time they called me at 6:00 in the morning, we had over 100
print jobs backed up in queues. By time I got in to work and determined that
it wouldn't be a quick fix, there were closer to 1000 print jobs in the
queues. You know how it is. The person prints something, doesn't see it come
out, so he prints it again... and again... and again.
I am wondering if anyone can tell why this happened. What was it about
upgrading the CP processor, which this server wasn't even using, and the
microcode that caused this Debian install to come screeching to a crawl? I
am worried about this because we are moving more and more servers to Debian
on our S390. We do upgrades at least every 3 years. In order to fix this
problem, I had to re-install my Debian server. Luckily, this was not a time
consuming process. We had the new server in about 3 hours, including 30 min
lunch and a 15 min break. Is this something I will have to do any time I
upgrade my S390? Any info would be helpful!
Sam Warren, Jr.
641-585-3535 ext. 7785