Re: Incoming feed statistics

In article <Pine.LNX.4.44.0305072141340.4175-100000@xxxxxxxxxxxxxxx>,
Jeffrey M. Vinocur <jeff@xxxxxxxxxx> wrote:
| I have this funny feeling there's a tool I should know of but don't...
| 
| We aim to carry a binary-clean feed, and for the most part I think we do a 
| pretty good job.  But lately I've been seeing some spikes in the incoming 
| traffic logs that make me wonder if something is getting past Cleanfeed.
| 
| It's not clear what the suspicious articles are, and I'm not sure how to 
| investigate them.  The obvious approach is to write some stats out with a 
| newsfeeds entry, but before I hack together something myself, surely such 
| a tool exists?

perl, perhaps. You can run a tiny script against your 'news' log and
find the large articles, or look at your unwanted.log to see if they
are coming in and being dropped. You can save your active file daily
and see if there's a spike in the new articles.

I don't think there's a canned tool, but you can get a lot of what you
need from just grep and one-screen perl programs (or awk, etc).

If you get feeds from a lot of hosts you can run feedcount daily and see
who has a spike and in what article size range, which might be helpful.

Is the spike in article count, byte count, or both?
-- 
bill davidsen <davidsen@xxxxxxx>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.