Sending traps vs. Hardware Status Polling

[fried1]

This came up at work today and wanted to give the most correct answer. Historically, our server were not configured to send SNMP traps to SIM; they relied solely upon Hardware Status Polling to be notified of any events. I don't believe that this will cover everything and I want to push to enable all servers to send traps to SIM. Can anyone comment on this? Are traps necessary to be alerted for system events (memory errors, hard drive errors, etc), or will the hardware status polling query cover everything?

Thanks,
Chris

[Albert Austin · Accepted solution]

SNMP is very important to make use of SIM's full potential. For monitoring and memory, harddrive erros you must have SNMP configured and running in your environment.
Hardware status is PING dependent.

[fried1]

To confirm, this is to say that any event other than up/down messages, as confirmed by hardware status polling pinging the machine, must be sent to SIM via an SNMP trap. Is this correct?

[Albert Austin]

Yes.

[fried1]

Thanks.

[Albert Austin]

If you follow the link below for a lot more info on SIM and how it works.

http://h18013.www1.hp.com/products/servers/management/hpsim/infolibrary.html

[Igor Karasik]

Chris,
>>To confirm, this is to say that any event other than up/down messages, as confirmed by hardware status polling pinging the machine, must be sent to SIM via an SNMP trap. Is this correct?

I think it is not exactly correct.
We don't use SNMP traps as well (don't configure traps destination at all),
and we RECEIVE all hardware alerts (drives, memory,....) from HPSIM.
Furthermore, we also use HP Proliant management pack for MOM and we receive all hardware alerts in MOM console as well.

So, it is interesting for me as well: which additional things we can take from SNMP traps?

[fried1]

Igor,

I'm interested to know how you're getting the information from the polling task? Are you using custom scripts or WMI?

Thanks.

[Igor Karasik]

Chris,
>>I'm interested to know how you're getting the information from the polling task? Are you using custom scripts or WMI?


Probably I don't understand/misunderstand your question, but we don't use not custom scripts and not WMI.
When hardware polling detect some problem (drive failure for example) server status in HPSIM change status from "normal" to "minor ("major"). We define in HPSIM smtp alerts for this status change and we received these alerts.

Actually, in our organization HP Proliant management pack for MOM send these alerts for MOM console as well, we use our MOM console as central console for all system alerts.