SMH and email alerts instead of SNMP traps

[dhermans]

Hi,

Using HPSMH on Solaris and Linux with SIM

Have SNMP all setup with read/write/trap community  strings, eg

trapcommunity mytrap
trapsink mysimIP

When I pull a NIC or power cable to test the system I get 3 emails to root with subject "Subject: HP Insight Management Agents Trap Alarm" and the event..BUT

the snmp trap goes to a machine I used ages ago to test SIM. It is not the trapsink. I've searched and searched and found no file with this old hostname or IP in it.

Ideas appreciated!! i can't find this config anywhere..

[Brad Cunningham]

Check the ILO configuration to see if its configured for SNMP pass-thru

As for the multiple alerts, I have see this with PSP 8.60 before for RH but have not dealt with Solaris much.

brad

[shocko]

Check for static routes/hosts entires also. Is the trapsink an IP or DNS name?

[dhermans]

Thanks - I checked iLO and it has the old SSO cert..

so maybe the O/S agents read the cert from the iLO to work out where to trap - bizarre

anyway working at last - Thanks

Now if only I could get rid of these endless traps:

(SNMP) Health Status Array Change occurred (11020)

3 boxes out of 150 constantly send this trap - and nothing wrong with the array..

SIM is good but so much work to make it useful:

- little niggles like above

- endless 11020 traps

- SSO to HP SMH no longer working

- SSO to iLO inconsistent

[shocko]

(SNMP) Health Status Array Change occurred (11020) relates to a bug in the insight management agents v 9.10. Is that what you are running? Simply change to v 9.00 ;)

[dhermans]

Thanks - i'll have to check how it is packaged

On Solaris it's:

 # pkginfo -l HPQhma
   PKGINST:  HPQhma
      NAME:  HP ProLiant Insight Management Agents
  CATEGORY:  system
      ARCH:  i386
   VERSION:  5.8.1
   BASEDIR:  /opt/HPQhma

and try downgrading...

Cheers