back up...

Ok.  Looks like we're back.  I would like to quote the posting i saw
somewhere over the past two days to the effect of "when disk space
runs out, all hell breaks loose". That's where most of the problems
stemmed from.

The issues are broken down below.  I think this was good motivation
(yaay, stable site, yaay!), but goddamn.  anyhoo, once avg. sleep
levels get back up above 6hrs a night, I'd like to give the helmet a
fresh install - having a more *cough*"standard" setup will help to defray any
of these hiccups (chokings?) in the future.

The issues were three fold: 

1) dns changed.

Diagnosis: new ip.  not sure if it was caused by below or not - just
as likely that the isp changed it.

Solution: install client to update ddns service w/ new ip's as they change.

2) / was full.  

Diagnosis: looks like the idiot sysadmin installed
apache and set it's home to be /www, instead of a more appropriate
/usr/local/apache/, /var/www/, or even /home/www. This is why httpd was not responding, even 

Solution: moved /www to /home/www; symlinked /www->/home/www.  berated the sysadmin.

3) /var was full

Diagnosis: caused by a extraordinarily large inbox (~80M).  on a
grander level, this should not have happened - I should
have set up mailbox quotas in the first place. originally thought it
was tomcat (is anyone using that service? d?), but now know better (i think).

Solution: user has been notified of bloated mailbox.  moved tomcat to
/home/tomcat4 (if this breaks anything, we can move it back), which
freed up trivial space comparatively. mail quotas are being set up.