[13106] in Athena Bugs
Re: OPERATIONAL BUG: root login w/o network
daemon@ATHENA.MIT.EDU (Richard Basch)
Thu Jan 19 07:27:17 1995
Date: Thu, 19 Jan 1995 07:27:11 -0500
To: mhbraun@MIT.EDU
Cc: "Richard Basch" <basch@MIT.EDU>, bugs@MIT.EDU, testers@MIT.EDU, op@MIT.EDU
In-Reply-To: mhbraun@MIT.EDU's message of Wed, 18 Jan 1995 17:33:24 -0500,
<199501182233.RAA22473@medic.MIT.EDU>
From: "Richard Basch" <basch@MIT.EDU>
From: mhbraun@MIT.EDU
Date: Wed, 18 Jan 1995 17:33:24 -0500
I actually had problems getting in to one of the AFS servers from the net.
I logged in doing an eklogin and it hung afte rthe 'motd' line until it
received an AFS client timeout.
My read on the situation was that we have that can be wedged by other servers
being down.
I am not sure what the timeout was for (as in why it cared).
Should we consider making AFSCLIENT False on servers ?
I propose that we add something to the crontabs that does a detach -O -a.
That way, if someone attaches a /srvd, for some reason, it will be gone
by the next day.
Or, maybe, we should add the detach to root's .logout.
We should also remove all the /srvd /srvd links that exist.
However, I also do not think that root should be timing out on things
like get_message. I have to see what else there is.
-Richard
