The general consensus was that it was a problem with the interconnect,
at least one suggested reseating cables and trying again. This wasn't
necessary, but we did need to bring down the other member of the cluster.
Everything worked properly upon bringing the other member down and
then booting both of them. The hardware was not touched. mc_diag and
mc_cable showed no problems.
Anybody know of a reason that the Memory Channel interconnect may have
entered an inconsistent state? Maybe a way to clear it up without downing
the member? (you know -- a way to avoid staying late so that the system can
be downed without affecting users etc.)
Charles Ballowe /"\
Unix System Administrator \ / ASCII Ribbon Campaign
cballowe_at_usg.com X Against HTML Mail
x3896 / \
Received on Fri Jun 21 2002 - 17:15:38 NZST