[theforum] only ~12 hours of downtime today!

David Kaufman david at gigawatt.com
Mon Nov 6 16:20:07 CST 2006


yippee.  or, more from the "glass-half-full" perspective:

50% more uptime, on average, than the lists have had all week! :-)

I kicked the server Sunday afternoon, it appeared to have been down 
since Saturday morning.  This after the downtime from Tuesday afternoon 
till Friday night.  I'm not sure if other sysadmins are kicking it too 
because the old kicks seem less and less effective, sometime the list 
mail starts flowing again after 30 minutes or an hour, other times not.

I've added to my list of things to do when the lists are bogged down:

 - truncate the clamav logfile:
   echo -n > /var/spool/qmailscan/qmail-queue.log

   (it was 70megs the other day, 19 megs just now)

 - purge the virus quarantine dir (was 35,000+ files in there...)

Legitimate list traffic is quite moderate.  It's the damned constant 
deluge of crap which spamassassin must analyze and filter, and the 
attached virii, which clamav must scan.

This is actually why I volunteered to take over for Dean when he dropped 
from exhaustion.  I wanted mailserver combat chops on a high-volume 
server with oodles of traffic, both legitimate and evil, so's I can sell 
m'self as an smtp warrior (when the perl coder market dries up, as it 
often it does).  So, I guess it's be-careful-what-you-wish-for time :-)

John and I are gonna put our collective noggins together Real Soon 
Now... hopefully tonight, if we can sync schedules... and tackle this 
problem and kick its butt.

You may expect to experience disruptions in service, time of 
instability, general mayhem and blood flowing in the gutters.

But Keep the faith, evolutionaries!  Though the enemy is legion, and 
we're armed only with the lowest-end Celeron money could er, rent ...in 
2003, our cause is Just and the open source Wizards are on our side :)

-dave (the lion-hearted) 




More information about the theforum mailing list