Servers with SP 8.00 show with false warning status now

[NJK-Work]

Hi

A coworker of mine recently installed SP 8.00 on some servers which are behind a FW. All of these servers now show an orange warning symbol in SIM. However, if you go to the SMH on the servers, they show all green.

I have tried removing the devices from SIM and re-adding, but they always come back as "warning".

Is there a port on the FW that we now need to open for the new 8.00 management agents that could be causing this? All other servers with 8.00 show as fine in SIM.

Thanks
Nelson

[Jimmy Rueedi]

Same at our site...
We also found the following:
The System Management HomePage doesn't show all of the categories anymore on those systems (e.g Operating System, Management Processor)

We didn't had time to completely drill out what doesn't appear anymore, so maybe it could be more.

Regards
Jimmy

[NJK-Work]

Thanks for the info.

Nelson

[NJK-Work]

The solution appears to be disabling WBEM as the default view in the SMH and re-enabled SNMP as the default view. Then two "warnings" show up in SMH. However, these warning (for CPU and Logical disk) are also false (but at least now they show up). I restart the HP agents and now the servers show clean in both SIM and SMH.

I tried just restarting the services without changing the SNMP/WBEM view in SMH first, but that does not work.

Nelson

[NJK-Work]

Need to disable WBEM view in SMH. Set it back to SNMP and then restart the HP agents and SMH services.

Nelson

[NJK-Work]

I actually found a new solution to this problem.

The source of this problems stems from the use of Thresholds (the Java based sliders for Disk, Server, and Processor in the SMH).

It appears the thresholds are not being reset properly when a server goes back to within normal operating parameters when running the 8.00 Management Agents. For example, if the process TH is set for 90% and the server hits 91%, it kicks off a warning idicator, but the server does not go back to "Normal" when the processor drops back below 90%.

This is componded by the use of the WBEM homepage view, which does not show this (it shows all normal even though SIM is reporting warnings). Switching to the SNMP view shows the "incorrect" errors, even though disk, server and process are all now running normally. Reseting the HP agents then resets this incorrect warning messages until the next time the thresholds become exceeded again, and then the whole process repeats.

The "fix" for this is to disable all thresholds (drag all the sliders to the far left, or 0 value).

Nelson

[Jarrod Kline]

I am having the same issue, specifically on x64 G5 hardware. We are using snmp. Once a threshold is met or exceeded a alert is generated, then upon returning to normal the status on the SMH never resets. Seems like the latest SMH fixes this issue on x32 2003 server systems, but not x64. Turning off the thresholds is not an acceptable fix to this problem. I called into HP and they claim a new fix will be out in the next PSP 8.1.x?

[Jerame]

Is there any status on this?

I recently started upgrading the PSP on my servers, and some of the post PSP files (the newer versions) and I am running into the same problem.

Does 8.1 PSP have a scheduled release?

[NJK-Work]

8.1 is now out. But I don't know if it fixes the problem.

Nelson

[lvinokur]

I recently installed PSP 8.10 on all servers, I have a similar issue.
I'm gonna have to check if the thresholds solve the problem and will get back with an update.
Leonid

[Mihai Hornet]

Hi,

I had the same situation with SIM 5.2 and PSP 8.15
This was because health status is checked by both, WBEM and SNMP.
If WBWM was selected as source for SMH all was green but when SNMP was selected it was orange because of iLO interface not connected. I disabled iLO monitoring in HP agents settings and HP SIM status showed green.

Mihai

[Davide De Michino]

Resolved!!!!
Thanks!!!