Mike... I'm gonna reverse course on this one.  I'm still digging deep into 
extended accounting, so I now realize that extended accounting is really 
useless on a system with a large number of long running processes.  The only 
way to get stats on long runing processes is via partial records via wracct... 
and doing that on hundreds of processes, especially when you don't know what 
they are, is really not a good solution.   

For instance, if I wanted to get extended accounting data for QA Zones running 
on a node that was useful, I'd need to do something like write out partial 
records for every process on the box every, lets say 24 hours... maybe 1 hour.  
Without doing that I wouldn't have any visibility into what was really going on.

I don't know... I don't like the idea of writing that kind of script.  Yuck.

There has got to be someone out there that is in fact doing this... I'd be 
interested to hear some feedback on it.

benr.
-- 
This message posted from opensolaris.org

Reply via email to