[555] in SIPB-AFS-requests
rosebud: problems
daemon@ATHENA.MIT.EDU (Richard Basch)
Mon Dec 2 21:25:16 1991
Date: Mon, 2 Dec 91 21:24:49 -0500
To: sipb-afsreq@MIT.EDU
From: "Richard Basch" <basch@MIT.EDU>
Today, I was informed of a few problems with rosebud.
Scenario:
- The following processes were dead: bosserver, vlserver
This might cause sluggish response if your primary vlserver is rosebud
(50-50 probability)
- The kernel receive queue for port 7007 (afsnanny) was moderately
large, and the port was bound, even though bosserver had died.
After a complete shutdown of all the related processes and a restart a
couple of times, things were consistent, but about a minute later, about
when the salvager completed, a message similar to the following
appeared.
rx_multi: out of memory
I don't remember the exact words, but that is the approximate message.
Anyway, by this time, after all the inexplicable behavior above, I
figured something was amiss and a reboot was probably the best course to
make sure that there would be continued stable service (the error had
not yet impacted on service, but probably would have led to problems in
the not too distant future).
That explains the 20 minute downtime of rosebud today. Most of it was
trying to reboot the machine (and a second attempt since the key was not
in Normal long enough, with a phone thrown in the trashcan for good
measure - I just wish I were around to have seen that part...)
-Richard
