Hello all -
I'm running Netview/NT 5.1 w/ SQL Server 6.5. I'm been grappling with a
problem for some time that i haven't found an answer to in the netview
archives.
Netmon will consistently die approx. 10 mins after a forced start; message
in netmon.trace is "insufficient available memory, disabling discovery."
Following line is "[netmon] connection to one or more services is no longer
available." NV.log reports "[trapd] an application has disconnected."
Furthermore, after force starting netmon, and giving it a few minutes to
restart discovery and node status checking, demand polling fails to operate
properly - it frequently stops at the message point "scheduled new node poll
at (time)..."
The problems appeared to begin some time after I disabled (for processor
load reasons) my scheduled ovtopofix and sql database cleanup entries.
They've since been re-started and run without a hitch.
Currently have 11,293 objects in my database. I've increased my ovwdb cache
size, run ovtopofix several times, dumped my sql database logs, adjusted my
polling speed to low, checked my polling intervals, and increased my nt
paging space by 25%, but the problems haven't improved in the least(although
my Netview/Solaris box keeps running like a champ.)
Any ideas from the user community? Have I finally run up against
Netview/NT's limits? Thanks in advance,
Oliver Grant
PEC Solutions (formerly Performance Engineering Corp.)
|