CIM7 alerting issues - hope Systems Insight Manager Alerting is better

[Anthony_67]

David

From Day 1 I have always had issues with CIM7SP2 alerting. We run CIM7SP2 on Windows 2000 Server. Initially I though my Event Queries were wrong but I called HP and set it up to their recommendation. I use the following:

Queries
All Queries
Event
All
Event(s) by cleared status: Status Not cleared
Event(s) of severity: Critical

All Devices

This rule gives me 5000 email alerts a week (no exageration). Ive tried so many queries, based on suggestions from this forum and from what Ive read. When the alerts do work, I usually get 20 emails for an event (most useless like Array Controller cache change). Do I really care when the Array Controller Cache changes??

Why can't CIM give me 1 email when my server goes down. I can get a free utility that pings a server and emails me 1 email when the server is no longer responding (I know this isn't ideal as the issue may not be server down, could be a link issue).. There is no rule that provides this kind of in depth information.

All I want from CIM are the following Alerts:
- Server Down, Server Up
- Memory Error
- Hard Disk Degraded, Hard Disk Failed
- Network Link Down, Network Link Up
- Power Suppy Failed
- Fan Failed

Just the std system checks - Ive posted so many messages in regards to alerting - Im praying it gets better in SIM. I know alot of other people here have similar issues with Alerting. If anyone can help me with my queries that would be fantastic

As a side note, I also have an issue on tbe main CIM screen where the MP status goes blue for most of my servers. The servers have the ILO working fine and the agents are loaded (I can connect to the ILO). We run SmartStart 6.3 and 6.4 on Windows 2000/2003

Thanks
Anthony

[Rob Buxton]

I'm not sure there will be a difference as it's the agents or other software on the Server sending the SNMP Traps that is the issue. All CIM or SIM does is process that information.

You need to work out what is generating the SNMP Traps on the Servers and then review that.
I just have queries that e-mail me on Critical or Major Events. I certainly don't get e-mails indicating an Array Controller Cache change.
As for iLO's, all I do is ensure the iLO's are included in the IP Discovery range. Discovery and in the ILO Config, entered the FQDN of the Server the ILO is in in the Insight Manager Integration section. Works fine for me.

[John Kramer]

In SIM I did notice that you can disable specific traps, so you could in theory disable all the traps you do not want to see.

Under Options>SNMP Trap Settings> Then you can select the MIB and Specific trap and enable or disable the trap.

[Jadrice Toussaint]

John -

That option was already part of IM 7 sp2.

[Scott Shaffer]

Anthony - my recommendation is to being excluding (using the exclusion criteria in IM7) events you don't want. So, if you don't want that cache change, go exclude that event type.

You could also include just the ones you wanted, but that can be hard because the events come from so many different subsystems with different traps (those from internal IDE drives are different from internal array drives, which are different from external SAN drives, etc.)

If you just want an email when a server goes down though - just create a query that looks for all the Critical Device Status Change events.

HP SIM does make it easier (in my opinion) to create email tasks because the setup is more intuitive and you are less likely to make an error.

[Peter Stringer]

If you are talking about getting smtp/eamil alerts then I had the same problem. For instance I needed to know about predictive failures on disks or even disk or raid failures. I was not getting these.

I found out that you had to ajust the severeties on the snmp page.

I think many snmp mibs are set to informational by default which will not get emailed.

I would go through all your important mibs and adjust the severity to either critical or major.

I did this and hey presto I got the emails I wanted.