Getting Email Alerts through SIM 4.2?

[S Campbell]

I'm brand new to HP SIM, but I did use the old CIM7. When using Compaq Insight Manager 7, I was able to create an email alerts when my servers' hardware status changed (eg. bad disk drive or power supply went out). Is there a way to do this in HP SIM 4.2? It seems that I can only set up an alert based on events not on server status. So far I've been unable to get SIM to email me even if I unplug a power supply (making the server's status = Major).

[Jason Hindson]

Try creating a rule based on Event Catagory. You may have to create multiple rules for the various subsystems you'd like to monitor (ie. Proliant Storage Events, etc.) but this has been working for me. We were particulary interested in receiving notification if a disk drive failure occurs. Since we use RAID it was not a critical event, but we still wanted to be notified when it occurred. So I had to establish a rule based upon event catagory to include status changes to storage subsystem.

[S Campbell]

Thanks much for your response Jason! It sounds like you are doing exacly what we are trying to do. We have Tivoli alert us when servers go down, but what we'd like is a warning that if have degraded hardware (particularly in our DR area). You mention "setting up a rule." How do you set up a rule? What I've been trying to do is set up a "Managed Task" (under Automatic Event Handling) that e-mails us if a disk goes bad. However it seems like there is no event created when I pull a drive out of an array on our test server. The server's status changes to "Major" but there isn't a new Event listed under "uncleared events." Do I have to configure something else to make these types of hardware failures trigger events?

[Jason Hindson · Accepted solution]

When I speak of rules, I'm refering to tasks in SIM. We are talking about the same thing, just using different terminology. As for the disk drive issue, I setup a task to alert me to any Proliant Storage Event. The event is triggered by an Event Catagory and then I selected from the rather large list, Proliant Storage Events. You should be able to accomplish the for other subsystems (memory, network, etc.). When I was testing I pulled one of the drives from my test server and the alert worked correctly. You may need to wait for a couple of minutes for the hardware polling to pick up the change. Give that a try and let me know how it goes. I've attached a picture of my rule in case I've been unclear.

[S Campbell]

Thanks again for your useful information! We are now using an alert much like the one that you showed. Our problem solution included one more step as well. We hadn't told our task to send an SNMP trap to our SIM server which meant that there was no events at SIM(thus, no email alert).

We now have a different problem. We are getting repeat events in SIM and corresponding emails. In our test we unplug a redundant power supply. The result is one set of events in the managed server's system event log but non-stop repeat events in SIM and Emails (every 30 seconds). The repeat events only stop if I disable the task and then re-enable. The events continue to be generated even after we replug in the power supply. Is there a way to only get one event and email when we lose a power supply? Or if this isn't possible, is there a way to reduce the frequency of SIM checks so that we only get a repeat event...say 1 every hour? I'm going to try to include our setup screens for the task so you can see how we set this up.

[Jason Hindson]

I'm not certain, but I believe you should remove the 'Forward as SNMP Trap' from your event definition.

You should configure the server/device itself to send the SNMP traps and not the SIM central management server. Try adjusting the SNMP agent settings on your test server with the proper Get/Set (Community Name) strings and then retry your test with the power supply. I don't think you want the CMS forwarding events as SNMP traps.

[S Campbell]

Booyah!! <--Industry standard technical terminology. Thanks much Jason! This fixed my problem. I think I changed both the SNMP settings as well as checked the 'Forward as SNMP trap' setting during the same test. So I thought adding that check box is what made our system receive alerts. Aparently our SNMP setting was to blame, not that particualar check box.

Thanks Again.

[itnas]

[quote]Or if this isn't possible, is there a way to reduce the frequency of SIM checks so that we only get a repeat event...say 1 every hour? I'm going to try to include our setup screens for the task so you can see how we set this up.[/quote]
Sorry but as for the last thing I think that during configuration of rules there is a "tick box" Clear Events.

After it setting we start to get event only once.

Sorry but we have other problem after this, we getting a lot of events from our cluster environment, for example if server is down we getingh message as from every virtual service of this cluster. Second about storage if there is something wrong with storage we get message from every server to witch this storage connected, in big environment it is alot of mails.