If you issue for example the NodeDown incident from the NNMi JMX console you should see the incident in "All Incidents" (I assume that dampening is not active for Node Down). AFAIR you must use the node name (not the FQDN or IP).
Following the documentation NO_TRAPD is correct, but I have seen new OM agent versions (11.1x), where NETSNMP is needed.
Yes, when I invoke incident with the right trap oid, I receive it in NNM All Incidents, but I don't get it in BSM OMi.
Probably the problem is not in the configuration, because the system worked fine for 2 years. But we had a power outage which affected the machine and I suspect that something went wrong when I started it. Unfortunately I had no idea what could it be.
As I said when I issue tcpdump on port 5162 I don't have any traffic, which suggest that I don't forward incidents to the Integration Adaptor.
I think we searched in the wrong place. I look through my bsmgw logs and I found that opr-gateway.log is full of [Thread-11571] ERROR SoapArgMsgList.endElement(104) - Sending events failure!
I did dbValidator.bat and it return failed:
### Index validation report ### ---------------------------------- Index check failed. Following issues found: - Missing index: indexOverCompT in table: bsm910_event.dbo.HISTORY_LINE in colum n: COMPONENT_TYPE. This column already has an index with different name!
Has anyone faced similar problems? I read that I have to use checkindices.bat tool, but I am not sure of the outcome.