On Tue, 26 May 1998, Mark Nordberg wrote: > Join the club. [...] > I can still do a df and see 99% full and out of space. It then starts > screwing up the server. This number started at 1000 and I kept increasing > it to see if made any difference, it doesn't. This is nntpcache 2.3.2 and > has had problems like this for a while.....i now have to set the time for > article expire to 2 or 3 days instead and not cache all I need. This is > on linux machine. Hmm, I'm seeing the same thing here.. I thought maybe I was the only one.. I've searched through the archives and have found some people who have 2.3.2.1 on Linux who claim to have it expiring.. I thought about emailing them to see what they were (or weren't) doing that I was missing.. I even tried downgrading to 2.3.2 which was worse, because other things broke.. Right now my expiration consists of doing a "killall nntpcache", then "rm -rf /news0/cache", then rm'ing the history and mmap files, then restarting nntpcache.. kinda sucks to have to do it that way though.. With our traffic here I'm having to do this once every 2-3 weeks with a 4 gig cache spool. If anyone out there has 2.3.2.1 expiring on a linux server how about posting to the list and let us know if you've done anything special? Shayne --------------------------------------------------------------------------- Shayne Hardesty | World Wide Web Page: System Administrator | http://sh.birmingham.al.us/ Quick Link, Inc. | E-mail: shayne@sh.birmingham.al.us --------------------------------------------------------------------------- "Trying is the first step towards failure." -- Homer Simpson ---------------------------------------------------------------------------