[1641] in SIPB-AFS-requests
Re: sipb cell outage this afternoon
daemon@ATHENA.MIT.EDU (mhpower@MIT.EDU)
Tue Nov 29 20:03:16 1994
From: mhpower@MIT.EDU
To: ghudson@MIT.EDU
Cc: sipb-afsreq@MIT.EDU
In-Reply-To: "[1638] in SIPB-AFS-requests"
Date: Tue, 29 Nov 94 20:01:39 EST
>I was doing this, ronald-ann went comatose. (The file server process
>currently has a core, so it may have dumped core.)
/usr/afs/logs/core.file.fs is dated November 23 at 09:43, and the only
entry in /usr/afs/logs/BosLog.old mentioning a coredump is:
Wed Nov 23 09:43:47 1994: Server process coredumped:
/usr/afs/logs/core.file.fs (signal 10)
I don't see much to connect this with a problem over 6 days later...
>I have no explanation for these events, and I find it very disturbing
>that unexplainable and (evidently) minor problems with the sipb
>servers have such a drastic effect on the Athena environment.
If no one understands this, and if the ronald-ann problem started
fairly soon after the shutdown of rosebud2, I really don't have any
confidence that there won't be similar outages after other steps in
the afs-server transition. Unless some explanation has come up since
this morning, I'dstrongly recommend against any further changes or
shutdowns to sipb-cell servers until the end of the term. There are
far too many people working against tight deadlines to risk any more
outages. And, after all, the rs6000 server isn't scheduled to fail
again until next summer :-)
Matt
