Re: leafnode creating multi-megabyte log files )-:

Am 16.05.2011 12:58, schrieb Arthur Marsh:
> Hi, I'm running leafnode from Debian unstable:
> 
> ii  leafnode                                             1.11.8-1
> 
> My config file is:
> 
> victoria:/etc/news/leafnode# grep -v \# config|uniq
> expire = 15
> 
> server = news.internode.on.net
> server = news.gmane.org
> only_groups_pcre = gmane\.
> server = news.mozilla.org
> only_groups_pcre = mozilla\.
> nodesc = 1
> 
> debugmode = 0
> 
> groupexpire comp.risks = 365
> groupexpire gmane.comp.graphics.opengraphics = 31
> groupexpire comp.protocols.kermit.misc = 365
> 
> maxcrosspost = 6
> 
> maxage = 15
> 
> filterfile = /etc/news/leafnode/filters
> 
> article_despite_filter = 1
> 
> hostname = ppp121-45-136-118.lns11.adl6.internode.on.net
> 
> and my filters file is:
> 
> victoria:/etc/news/leafnode# more filters
> ^Newsgroups:.*[, ]gmane.spam.detected$
> ^Newsgroups:.*[, ]gmane.spam.detected,
> 
> I'm also running logcheck and it is spending hours processing log files
> of the form:
> 
> May 16 14:05:59 victoria fetchnews[24378]: gmane.test: killed 3423
> (<el9g92$uv5$
> [email protected]>), too old (1621 > 15) days
> May 16 14:05:59 victoria fetchnews[24378]: gmane.test: killed 3424
> (<ela7ta$7g0$
> [email protected]>), too old (1621 > 15) days
> 
> Is there any straightforward way to stop leafnode writing such log files
> to begin with?

You haven't shown the fetchnews command line (from shell or cron) that
triggers this kind of logging.  The -x option is notorious for that,
naturally :)

Also, if your fetchnews runs do not complete successfully or cannot
update (= write to) the last-fetched records in the server-status files
(.../leaf.node/news.gmane.org), that situation can cause such logging.