webalizer stopped working

blasty

Verified User
Joined
Feb 28, 2008
Messages
17
Heya,

A week or what ago I 'downgraded' from a custombuild to a customapache setup for various reasons which aren't of any concern to this thread. Anyway, everything worked out fine after some tweaking here and there. Except there's one thing that *somehow* stopped working, webalizer.

I've tried to execute the webalizer binary from a shell to see if the binary works at all, which it does. (since it couldnt compile webalizer i had to grab a static binary somewhere)

As far as my knowledge goes, the tally cron is supposed to regenerate the webalizer stats. That cronjob is still there and tally is getting run every 24 hours. Adding the tally to the task.queue manually didn't solve the problem either, so the problem is definitly *not* the cronjob.

The weird thing is, AFAIR the webalizer binary was never touched during my custombuild -> customapache process, so im kinda confused why it stopped working.

My question: is there any way to get a log report of the tally process to see what might have gone wrong there? Or does anyone recall a similar situation and knows about a fix?

All help is greatly appreciated.

Kind regards,
blasty


P.S. Some additional system info, for what it's worth:
www1# uname -a
FreeBSD www1.xxx.xx 6.3-RELEASE-p1 FreeBSD 6.3-RELEASE-p1 #0: Fri Feb 29 02:14:22 UTC 2008 [email protected]:/usr/obj/usr/src/sys/QUOTA i386
www1# webalizer --version
Webalizer V2.01-10 (FreeBSD 6.3-RELEASE-p1) English
Copyright 1997-2001 by Bradford L. Barrett
 
Ok. I should've checked out /var/log/directadmin/ way earlier. Some more relevant stuff:
Code:
2008:04:02-02:30:06: Unable to run webalizer for xxxx.xx: Warning: Truncating oversized hostname
Skipping bad record (1)
Warning: Truncating oversized hostname
Skipping bad record (2)

And this basically happens for *all* domains, yet the logs in /var/log/httpd/domains/* keep growing. I think the DA dataskq is a bit confused here. But there's not much I can do, now is there?

Edit:
More investigation done. The cause doesn't seem to be webalizer but apache that is spitting out crippled logs. There's isnt any real log entries in any of the domain's access logs, instead every single line says 'combined' (which is the log format which should be parsed there, not just the word written dammit.)

Any ideas?
 
Last edited:
One step at a time, I progress through this maze.

I just figured out why apache wasn't writing the proper logentries, for people who wonder, I had the LogFormat entries *inside* the <IfModule log_config_module> directive. When I put them outside of this directive the logs started working again.

YET, running a tally does *not* update the webalizer stats. :(
When I run webalizer *manually* it _does_ update the stats. Any ideas?
 
Solved. Not too sure what I did to make it working in the end, probably more patience from my side was required. Too much coffee. Have a nice day!


Peace,
blasty
 
Back
Top