[leafnode-list] Killed count in log incorrect?

[Laurence]

On Monday, January 19, 2004 3:16 AM, Matthias Andree
<matthias.andree at gmx.de> wrote:

> Laurence schrieb am 2004-01-17:
>
>> Hi, I'm running version 1.9.43.rel.  I've set up a few entries in
>> filters
>
> Please update to 1.9.49 lest you want others left able to lock up
> fetchnews.

Ok, I finally got round to upgrading...

>> and I'm 100% sure that articles are matching and being killed.
>> However running:
>>
>> grep killed fetchnews.log | grep -v " 0 killed"
>
> Unfortunately, filter logging is not uniform, so this won't do.
>
> Try: egrep '[kK]illed|[dD]iscarding' fetchnews.log
>
> Note this is logged with priority info and facility news. Make sure
> your syslog.conf settings actually capture this - something like
> "news.notice" will not catch "info" logging. You may then also want to
> grep -v 'already fetched before' if using multiple upstream servers
> with overlapping news groups.

With 1.9.49 I'm seeing non-zero killed counts, but these seem to apply to
articles that are killed because they have been already loaded from other
servers.  Is there a way to see how many articles are discarded because they
match a filter?

Thanks,
Laurence