david@bdt.UUCP (David Beckemeyer) (04/05/89)
I have a stupid question that I have never seen a complete answer to
although it's probably right under my nose.
I am confused about some of the temp files that rnews creates when delaing
with incoming news. I haven't found a description of it in my news docs.
I've been running news for a little over a year and the on-going problem
has been always running out of disk space. I have a single filesystem of
about 34 Mbytes dedicated to /usr/spool/news. I expire most groups after
5 days and others sooner. I keep 21 days of history.
Anyway, back to my question. During receipt of news (I think that's when
it happens), rnews will sometimes create temp files (I guess they're temp
files). These are sometimes files with long numeric names, such as
8903231203109a. It sometimes puts the files in /usr/spool/news, but it
also puts them in /usr/spool/news/.rnews. It also creates files of the
form .ara01190 and .ina01190 in /usr/spool/news. My question is what are
these files and what does each type contain? More importantly, where is
all this documented besides in the source code?
My (poor) understanding is that these files are created when rnews decides
that it cannot save the articles right now. Is this correct? How does
it decide this?
--
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 Morajbuck@epimass.EPI.COM (Joe Buck) (04/06/89)
In article <537@bdt.UUCP> david@bdt.UUCP (David Beckemeyer) writes: >I am confused about some of the temp files that rnews creates when delaing >with incoming news. I haven't found a description of it in my news docs. The files of the form .arXXXXX and .inXXXXX are temporary files made by rnews when it installs an article. One is the article without the header, and one is the article with the header. They never should be left around, but frequently are when error conditions arise, like "disk full". They aren't documented because they aren't supposed to be there. (Famous last words). The files with long numeric names in the .rnews directory, on the other hand, ARE documented. If expire is running when articles arrive, or if SPOOLNEWS is defined, incoming news isn't processed until an "rnews -U" is run (expire forks one of these off when it completes). These files are usually compressed batches, but they may also be individual articles (if you get some news unbatched, or if a local user posts). "rnews -U" digests all the stuff in the /usr/spool/news/.rnews directory; you can run this if you like if you end up with files there unexpectedly. -- -- Joe Buck jbuck@epimass.epi.com, uunet!epimass.epi.com!jbuck