Rich Shepard wrote: > On Wed, 20 May 2009, Roderick A. Anderson wrote: > >> I admin 10+ (ran out of fingers and my shoes are on) mail servers all >> running pflogsumm and the shell script does not futz with pflogsumm.pl. >> If for some reason Date::Calc or any other module wasn't installed >> during the OS installation it got installed. > > Rod, > > The docs mention that Date::Calc is required and that most current > distributions include it. I think the check is like our appendix: left over > from earlier evolutionary times and no longer necessary. > >> I'm going to have to go look at the source distribution of Pflogsumm as I >> run CentOS and pflogsumm comes from a RPM package. Maybe that's why I've >> never had problems even when upgrading CentOS or coming from a Fedora >> install. > > I've not had problems until a month or so ago. I've used pflogsumm since > 1998 (or 1999) and received a mailed report each day. Why it stopped its > daily run is what I'm trying to find.
Was there anything else updated/upgraded around that time. I remember some of that original thread but there didn't seem to be. Being kind of pragmatic I'd drop the problem script and either write my own or use the cronjob examples in the script possibly with those extra OPTIONS in 1pflogsumm. One thing I've had done is run the report just after midnight for the previous day using the '-d yesterday' switch. Let's me know what happened the day before and watch for trends. But there are scripts that run on each of the mail servers that monitor the queues and send an alert via a SMS gateway when they exceed some set level. Good luck with find what is causing the problem. There are days I _really_ hate computers! Rod -- > > Thanks, > > Rich > _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
