[4568] in Athena Bugs
[Patrick J. LoPresti: AFS and living groups]
daemon@ATHENA.MIT.EDU (Richard Basch)
Mon Mar 19 20:42:08 1990
Date: Mon, 19 Mar 90 20:41:47 -0500
To: bugs@ATHENA.MIT.EDU
From: Richard Basch <probe@MIT.EDU>
(sending to bugs for proper demuxing... I have not forwarded this
message elsewhere... I also have no idea how he knew my involvement with
AFS, if he did - I certainly don't know who he is, and I am getting
rather tired of getting people referred to me without my consent,
especially since I already have a full plate).
"r" is the old rpc that was used; we are now using "rx", but it probably
has similar timeouts - I know it doesn't work on low-speed networks, but
unfortunately, fixing the timeouts on one end will probably not suffice,
and I do not know the implications of what would happen if we talked to
an outside party with a different set of timeouts...
-Richard
------- Forwarded Message
From: patl@ATHENA.MIT.EDU
To: probe@ATHENA.MIT.EDU
Subject: AFS and living groups
Date: Mon, 19 Mar 90 17:52:39 EST
Hello!
I am in charge of the Athena cluster at Zeta Beta Tau
fraternity. We are connected to campus by a 14.4Kbaud line, which
appears to be too slow to support AFS use. In particular, even one
client running AFS seems to be too many, as it pins the modem,
panics, and reboots whenever I try to access a file of any reasonable
size.
It seems plausible to me that this problem is caused by some
timeout being too short for our modem to support. From campus, I
attached afsdev and tried to make some sense out of the source. I am
no kernel hacker, but in the R package (/mit/afsdev/vax/r) I see that
retransmit timeouts of 2 seconds are set in both r.c and rftp.c. Is
it possible that this is the source of the problem? If not, what
might be? More importantly, is there anything I can do to get AFS
running over here?
Thanks very much for your time,
Pat
------- End Forwarded Message
