[8321] in Hotline Meeting
master timed on 18.70 subnet, and possible problem.
daemon@ATHENA.MIT.EDU (Calvin Clark)
Sat Mar 28 16:41:06 1992
Date: Sat, 28 Mar 92 16:40:36 -0500
From: Calvin Clark <ckclark@mit.edu>
To: hotline@Athena.MIT.EDU
Cc: network@Athena.MIT.EDU
Reply-To: ckclark@mit.edu
In-Reply-To: [8318] in hotline
Short message: prill's timed running okay...problem solved
Long-winded message:
I caught Brian in the machine room and got him to look at prill, because
the master timed on the machine was not "mastering", i.e., `timedc m' on
any machine on the 18.70 subnet returned "communication error"; I had
started a master timed on podge (See [8318] in hotline) as a temporary
measure, since the drift on many of the machines was great enough to
cause Kerberos "time out of bounds" errors. When he logged in, prill
was *running* /etc/athena/timed -M, but this obviously wasn't working,
because when I made the timed on podge a slave again, there was still no
master on the subnet. Killing prill's timed process and restarting it
seemed to solve the problem. There were no errors logged by timed in
/usr/adm/messages. Things seem fine now.
-Calvin
