Good evening.
I have some sorrowful news. Our shell server, sekhmet
(sekhmet.lug.mtu.edu), has broken in such a way (see Technical Details
below) that we are unable to repair it without ordering a new
motherboard.
However, we have realized that we have a perfectly good, if
power-hungry, machine already in the rack, which could easily be brought
online as a replacement. Thus, this evening Matt and I have brought
imhotep (an IBM NetFinity 5600R, 2X 866MHz Xeons and 3GB RAM) back to
life, fully up to date and online.
One caveat is that all of the data on sekhmet remains on the two SCA
drives that sekhmet used. Since we do not have a safe port to plug them
into to migrate the data, Matt will be doing that tonight and tomorrow
at the Nerdery, and we will image imhotep Monday or Tuesday night.
Therefore, please expect sekhmet.lug.mtu.edu to refuse your connections
until otherwise notified. Much apologies for the downtime, but we are
working hard to resolve it.
TECHNICAL DETAILS:
sekhmet is an HP Proliant DL360 G3 machine. During our maintenance
session this evening, dedicated to putting in a RAID card to allow
sekhmet access to the IBM drive shelves, sekhmet began failing POST with
a 1611 error code. This indicates a failed CPU fan assembly.
However, the fans all appear to be spinning normally, and after
reseating them all and the board as a whole, talking with HP support and
reseating processors and PCI cards, we were informed that there were no
other solutions but to replace the motherboard. We felt this was
extreme, so Matt and I looked up further information online, leading us
to attempt clearing the NVRAM. This approach did not work, however, and
so we have run out of ideas.