Systems Management (OpenView-OP Mgmt) Practitioners Forum

Agent hearthbeat message issue

Go to solution
Outstanding Contributor.

Agent hearthbeat message issue




i have a question to experts :)

OMW 8.16

Sitescope 10 & 11


My architecture:

12 sitescope that forwarding alert to OMW via opcmsg command. So except this 12 servers, i haven't any others managed nodes recorded in my database. All nodes monitored by SIS are considered as external nodes.


yesterday, my msgqueue on manager grows up to 50Mo, delaying processing treatment (i received a little event storm and 10h later queue was at 20Mo). When i activate the "check before managed nodes" option in external nodes, my queue drop to 0 in 3minutes. I thought that name resolution is the root cause of my incident, because as i know, external nodes feature disable the name resolution.

Problem, now OVO internal message from agent hearthbeat are treated as external node, and the source node field display something like {a65-fe.....} and messages no more appear in managed node browser.


I try to change my external node pattern from <*> to [!managed_node*] and messages don't appear (all of them)!! As all my alerts come from the same nodes (the SIS), all alert match the pattern (i think, maybe i'm wrong), because pattern seem to be based on node sender and not retrieve from node argument passed to opcmsg command...


Many many points to every one could help me to find a trail or to solve and understanding this issue ;)


Outstanding Contributor.

Re: Agent hearthbeat message issue



i finally change the external node parttern matching by one of this form:

<![{Unique node ID}]>


Now it's working.