I opened up a case with HP on this issue and this is the reponse that I got from the support team. And I must admit, I don't know how much HP is paying these support guys but I was very unhappy with the way this case was handled. The support guy was basically saying anything and did not want to do any research but instead was looking for easy way out to end the call.
My issue:
Prior to IM7 SP2, I was able to receive both warning events and critical events from IM. Because the warning threshold events were not of significance to me, I disabled the warning thresholds on all servers monitored by IM. When I upgraded to IM7 Sp2, the warning thresholds now registered in IM7 SP2 as critical events not as minor which was the case prior to IM7 SP2. Futhermore, because I disabled the warning thresholds, everytime I rebooted a server, the following trap is sent to the IM7 SP2 server:
Threshold Agent: Rising Threshold Passed. SNMP MIB Variable .1.3.6.1.4.1.232.11.2.4.1.1.5.3 has value 0 >= 0. Severity: Warning Description: T:SQLLog [NTFS] disk space. Refer to the MIB for a definition of the variable. [SNMP TRAP: 10005 in CPQTHRSH.MIB]
at first I was not sure as to why this is happening so I opened up a case with HP and spoke to a tech who suggested that I uninstall and re-installed the Management agents (version 6.40). This did not solve the issue. I then followed up later and my call was escalated to another tech. This guy needs to get off his high horse and do what he's paid to do. He refused to try anything that I suggested but instead was ftrying to feed my a lot of BS as to why this is happening and there was nothing that could be done and I should accept the way it is.
This is an email from the first tech guy about this issue:
Jadrice,
I have inquired into the warning and critical threshold traps question from IM7 development and unfortunately this is as designed.
Basically a decision was made based upon feedback from many of our other customers to limit the number of traps that would be generated by both the agents and forwarded by IM7.
Apparently there was concern from many of our customers over our agents being too "chatty" and generating more traps than they thought necessary. A decision was made to limit them wherever we could and in your case there will be only one critical trap generated for both the warning and critical disk threshold alarm.
I know this is not the response you wanted to hear but it is the truth. I did make a point to express your frustration as to this behavior and the points were documented and will be taken into account for future enhancements to the product.
In the meantime the only advice I have is to edit your queries and notifications to adjust to this behavior.
I apologize that there will not be an immediate adjustment to this behavior.
If this is a business decision, how can HP overlook the critical portion of this issue. If customers were complaining about agents being too chatty, the issue should have been eliminated form the agents side and not IM7 since teh traps are beng sent from the management agents. So to me this is BS.
Here is the email from the second tech that swore that I didn't know what the hell that I'm talking about:
Mr. Jadrice Toussaint
Your case has been evaluated by the ProLiant Essentials Team and these are the solutions to the issue encountered :
For a critical threshold to be set at 95%, a valid range for warning threshold is 6 percentage to 95 percentage. If this range is set below this point as shown on number 1 screen shot event viewer generate the warning message and IM 7 will report a critical event. Currently IM 7 is using the same message for the Warning and Critical setting. This has been sent to our AET personnel for a suggested improvement to IM 7.
Well, why would I need to set up a warning threshold if it is being reported in IM7 not as a minor event but as a critical event?
The way I configure my alerting, I use a product called ATTENTION that dials out and call the appropriate member of my team when a critical event is received in IM7. Needless to say that I do not want to send a false positive notification in the case of disk space threshold. So this is unacceptable.
I'm suppose to be using an enterprise based product (IM7 SP2)that was supposedly tested thoroughly before it was released.
Come on guys, you need to fix the agents that are sending the traps not IM7. Instead it appears to me that you broke the way IM7 SP2 is hadnling disk space threshold events.
If you want to eliminate warning thresholds, then take out of the management agents nto IM 7 since it only listens to SNMP traps.
