testing that snmp is set up correctly

[Scott Sievers]

We recently had a server experience a hardware failure and the server didn't notify Insight Manager (CMS). The problem was the community string and destination were not set up in the SNMP settings. This has been fixed.
Now I need to prove that the rest of our servers can send out alerts to the CMS. I could log in to each server and send a generic test trap to the CMS, but this would take too much time. Is there any other way I can verify that every server can talk to the CMS?
Is there anything I can get Microsoft Systems Management Server to execute from each servers?

[Stacy Rogers]

You could try trapgen.exe. That would let you send the test trap from the command line.

trapgen -d destination -c public -i sourceIP -o 1.3.6.1.4.1.232 -g 6 -s 11003

[Scott Sievers]

I tried TrapGen, but it bypasses the snmp service on the server. It uses its own community string setting and destination setting to send the snmp trap. I need to send a test trap from the server to the CMS and use the SNMP service on that server. Is this possible or do I have to manually click "Send test trap" on each server?
thanks

[Stacy Rogers · Accepted solution]

The best I can think of is running a Set Disk Thresholds task through HPSIM and setting a dummy threshold on disk space. I think the threshold has to be at least 6%. Then when the agents poll, you'll get the low disk space traps from those configured correctly. Afterwards, you can remove the dummy value.

[Stacy Rogers]

I found out something else. You could do an snmpset to the cpqHoGenericData string and that would generate the test trap.

For example: "snmpset -c public servername 1.3.6.1.4.1.232.11.2.8.1.0 octetstring TESTTRAP"

Of course, you'll have to know the set community string on the systems.

[Scott Sievers]

I think snmpset would also bypass the SNMP service running on the server. Then I wouldn't know if the SNMP settings were set up correctly or not. I do, however, love the idea of setting the disk threshold level to 6% for every server. Any server that doesn't send me a low disk space warning isn't properly configured. That's brilliant! Problem resolved

[Brett Boone_1]

You can also send a "test trap" from the management agent applet under control panel. This will send a test msg to the console; and email if setup; without having to set any values on the server. Just food for thought.

-BB

[Scott Sievers]

But the issue with the control panel app is that I have to manually log into each server and send the trap. I need a way to make every server automatically send me a test trap. For now the only option we have is to reset every threshold to 6% and then see which servers send out alerts.

[Raymon_1]

I was hoping for a HP-SIM7 task for this, bacause i`m having the same problem.

I just need a Cimphonehome task in my HP-SIM System

[Pedro VERA DE LA ROCHA]

Hi all.

I am very interested on this conversation, I am also trying to test my system, where several HP Proliant servers have HP Insight Manager Agents installed (on the top of Red Hat Linux).

However, I don't use a Network Managament Station with HP SIM, instead I have an XP PC running MIB Browser which can manage and receive SNMP traps from all systems.

Still, I cannot make to generate traps at the managed servers, if not manually by using the tool 'snmptrap'.

If I shut down an interface with 'ifconfig eth1 down', no trap is generated.

Nor I can use 'snmpset -c public servername 1.3.6.1.4.1.232.11.2.8.1.0 octetstring TESTTRAP' because I don't have the corresponding CPQHOST-MIB loaded on the server.

And setting disk thresholds is not possible since I am not using HP SIM, unless I know what MIB variable to set, and I cannot figure out.

So, please, if you could give me some hint...

Thanks.