Service Guard SNMP trap sending

[Quark]

Hi,

we want to monitor our Service Guard clusters through SNMP. there is apparently a dedicated snmp sub-agent that runs for Service Guard.

The problem that we are having is that we get very few SNMP traps. I did some network tracing (using nettl) and saw the I only receive 2 snmp traps when a package is halted. That's it. When a package gets started, no traps.

I do not know much about Service Guard but can you config SG to send more traps?

Kris

[John Bigg]

You cannot configure the traps that Serviceguard sends out but you should get more than this. You should see the following:

hpmcSGSubagentUp Signals the management station that the ServiceGuard subagent is up
hpmcSGReconfig Cluster is in the reconfiguration state
hpmcSGClusterUp Cluster is up
hpmcSGClusterDown Cluster is down (see section 6.1.4)
hpmcSGConfigChange Cluster/package configuration has been applied (via cmapplyconf)
hpmcSGPkgStart Package has started
hpmcSGPkgUp Package is running
hpmcSGPkgHalt Package is halting
hpmcSGPkgDown Package is down (unowned)
hpmcSGSvcDown Service is down
hpmcSGPkgFlags Packageâ s global or local switch enable flag has been changed
hpmcSGIPAddrUp MC/ServiceGuard added a relocatable address
hpmcSGIPAddrDown MC/ServiceGuard removes a relocatable IP address
hpmcSGLocalSwitch MC/ServiceGuard has completed a local switch from one network interface to another
hpmcSGSubnetUp MC/ServiceGuard detects a previously disabled subnet is now operational
hpmcSGSubnetDown MC/ServiceGuard detects a previously operational subnet is now down
hpmcSGNodeUp Node joins active cluster membership, could be after cluster reforms
hpmcSGNodeHalted Node has been removed from active cluster membership as a result of an operator-initialed command
hpmcSGNodeFailed Node state has changed to failed