1.9.50.rel: "groupinfo file not present or too small"
Brought to you by:
m-a
Menu
▾
▴
#5 1.9.50.rel: "groupinfo file not present or too small"
With 1.9.50rel, I keep getting mails from my fetchnews cronjobs telling me "Groupinfo file /var/spool/news/leaf.node/groupinfo not present or too small, forcing active fetch.". Running 'fetchnews -f' manually didn't help. Here's the command's output, maybe it helps - it does contain one warning message, but I don't know if it is related to this problem:
[root@airframe]:/var/spool/news# su - news -c "/usr/local/sbin/fetchnews -v -f"
leafnode 1.9.50.rel: verbosity level is 1, debugmode is 0
Forced active fetch requested from command-line (option -f).
Groupinfo file /var/spool/news/leaf.node/groupinfo not present or too small, forcing active fetch.
Trying to connect to news.individual.de:nntp ... connected.
Using STAT <message-ID> command on this server.
news.individual.de: 0 articles posted
Getting all newsgroups from news.individual.de
Disconnected from news.individual.de.
Trying to connect to news.telekabel.at:nntp ... connected.
Using STAT <message-ID> command on this server.
news.telekabel.at: 0 articles posted
Getting all newsgroups from news.telekabel.at
Warning: skipping group "max.", invalid name (NULL component).
Disconnected from news.telekabel.at.
wrote active file with 47398 lines
Started process to update overview data in the background.
Network activity has finished.
Running in very, very, very, very vebose mode didn't shed any light, either:
[root@airframe]:/var/spool/news/leaf.node# su - news -c "/usr/local/sbin/fetchnews -vvvv"
leafnode 1.9.50.rel: verbosity level is 4, debugmode is 0
Groupinfo file /var/spool/news/leaf.node/groupinfo not present or too small, forcing active fetch.
Trying to connect to news.individual.de:nntp ... connected.
Using STAT <message-ID> command on this server.
news.individual.de: 0 articles posted
Getting all newsgroups from news.individual.de
Reading server info from /var/spool/news/leaf.node/news.individual.de
[...]
The groupinfo file does exist, and it has the proper permissions, AFAICT (at least previous leafnode versions didn't complain):
[root@airframe]:/var/spool/news/leaf.node# ls -la
total 2624
drwxr-xr-x 2 news news 4096 Feb 19 20:40 .
drwsr-xr-x 36 news news 4096 Feb 19 20:45 ..
-rw-rw-r-- 1 news news 2664234 Feb 19 20:40 groupinfo
-rw-rw-r-- 1 news news 1307 Feb 19 20:39 news.individual.de
-rw-rw-r-- 1 news news 1291 Feb 19 20:39 news.telekabel.at
I have no idea if it is "too small" (?) - leafnode is supposed to care of this by itself, no? Anyway, diskspace is not an issue:
[root@airframe]:/var/spool/news/leaf.node# df /var
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/hda5 1032056 344080 635540 36% /var
[root@airframe]:/var/spool/news/leaf.node# df -i /var
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/hda5 131328 30343 100985 24% /var
Any hints or ideas what is causing this message?
Logged In: YES
user_id=2788
Wow. This one escaped the release candidate tests on Linux
and FreeBSD, and I cannot figure the condition that might
cause it here.
One error check is bogus and inherits an "ENOENT" that a
previous, unrelated to the groupinfo load operation, system
function left in the "errno" variable. I am attaching a
patch - please apply and let me know if it fixes your problem.
cleaned up patch to fix bogus "groupinfo too small" message
Logged In: YES
user_id=2788
You can also try leafnode 1.9.51 instead.
Please attach a comment that states whether 1.9.51 or the
attached patch on top of 1.9.50 fix your problem.
Merci d'avance.
Logged In: YES
user_id=150806
Yes, that did it. No more "not present or too small" messages. Thanks!
Logged In: YES
user_id=150806
Ooops, your second comment was added while I posted my answer. 1.9.50.rel plus your patch works fine, will try 1.9.51.rel soon.
Logged In: YES
user_id=150806
Just FYI, 1.9.51.rel works fine, too.