david@bdt.UUCP (David Beckemeyer) (05/03/89)
I know I'm re-hashing an old story but I'd still like to hear from
others who have solved these problems.
The Environment:
A) My system is running Microport System V.2 so it has the "inode bug".
B) I have 35 MB for /usr/spool/news, as a separate file-system.
C) I have about 6 MB free for /usr/spool/uucp and other /usr stuff.
D) I may have as little as 1 MB free for /tmp (part of / filesystem)
E) I have two outgoing feeds (all using UUCP).
The Problem:
Often (A) above causes me to lose news batches when there are "really" many
megabytes of disk space left. Rnews just won't quit. It keeps trying
and failing to receive the batch(es) causing lot's of lost articles.
It seems that if rnews would quit or if there was a hook for checking free
space before processing the next batch, it would help. Does anybody
use such a hook, perhaps enveloped around rnews?
I saw during the "inode bug" discussions a while back that some people
run fsck via cron to fix it. How does one "cleanly" do this on a
multiuser machine?
I have a sendbatch front-end which checks space in /usr/spool/uucp
before queueing more files which "solves" problems related to (C) and
(E) above and would mean that the incoming articles could sit in
/usr/spool/uucp for a little while if they had to.
There seems to be no problems caused by (D) above in my configuration.
This discussion is probably "old news" to most of the net, so replies
by e-mail are probably best.
--
David Beckemeyer (david@bdt.UUCP) | "Adios amigos. And, as they say when
Beckemeyer Development Tools | the boys are scratching the bad ones,
478 Santa Clara Ave. Oakland, CA 94610 | 'Stay a long time, Cowboy!'"
UUCP: {uunet,ucbvax}!unisoft!bdt!david | - Jo Mora