[leafnode-list] "cannot obtain group list" error running texpire

["William Grinolds" <wgrinolds@xxxxxxxxxxx>]

Okay, I'm new to this mailing list, I've searched the archives, but I
couldn't find a reference to this issue.

Here's the scoop: Leafnode has been running beautifully for over a year,
no problems through several different revisions of 1.9.x.  I've even run
the latest version, 1.9.24.rel, since shortly after it was released.
However, at some point, texpire stopped working.  This was well after
1.9.24.rel was installed and functioning properly for several weeks.
When I run "texpire -f" every morning at 04:00 (via cron), it now no
longer expires articles, making my disk fill up.  Fetchnews and leafnode
are working properly, with no problems.  When I run "texpire -f"
manually, the first thing it says is "cannot obtain group list",
followed by a blurb saying "total: 0 articles deleted, xxxx kept".
Adding -vvvv gives no additional relevant data (it simply adds "texpire
1.9.24.rel: check mtime only").  The log file gives no clues as to what
file it's specifically trying to access but can't, as shown here (these
are the ONLY two lines it adds):

Sep  9 18:09:26 grinolds texpire[4811]: wrote active file with 28972
lines
Sep  9 18:09:45 grinolds texpire[4811]: 0 articles deleted, 5853 kept

The only thing that I can safely say that might have been a factor is
the fact that I moved my /var/spool/news partition and expanded it from
2 gigs to 2.6 gigs (yeah, big change there).  The next morning at 04:00,
the problem started.  I'm thinking I may have some permissions set wrong
somewhere, but I don't see it.  Permissions on /var/spool/news is
drwxrwsr-x and is owned by both user and group news (just as the INSTALL
had specified).  Fetchnews et. al. created all subfolders (with the sole
exception of "message.id", which I had to create myself, and fetchnews
reset permissions on that subfolder by itself to drwxrwsr-x, with news
as both user and group owner), so I don't see how that would affect it.
Permissions on /etc/leafnode/options are -rw-r-----, owned by user root
and group news, so there's no perceivable issue there (otherwise, why
would fetchnews and leafnode run?).  I've even cleared out the partition
(next best thing to reformatting it) and started over several times, but
that didn't change the situation.

Here's the output from when that new partition (2.6 gigs) was formatted:

mke2fs 1.27 (8-Mar-2002)
Filesystem label=
OS type: Linux
Block size=1024 (log=0)
Fragment size=1024 (log=0)
653312 inodes, 2610531 blocks
130526 blocks (5.00%) reserved for the super user
First data block=1
319 block groups
8192 blocks per group, 8192 fragments per group
2048 inodes per group
Superblock backups stored on blocks:
        8193, 24577, 40961, 57345, 73729, 204801, 221185, 401409,
663553,
        1024001, 1990657

I'm pretty sure I have enough inodes (when the partition did fill up, it
still had more than plenty left).

Can anyone help point me in the right direction?  BTW, I run Slackware
9.0 beta 1 (which is simply Slackware 8.1 plus a small handful of
modified/upgraded packages), Linux kernel 2.4.18 + freeswan 1.98b mods,
gcc 3.2.  If you need any other info, please let me know.

Thanks for any help!

Bill


-- 
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx -- mailing list for leafnode
To unsubscribe, send mail with "unsubscribe" in the subject to the list