Hi all.
I have one server VMware ESX 4.0.0 Build 208167 attached on a vCenter Server 4.0.0 Build 208111.
Esx is installed on a Cisco UCS Blade and all seem to work fine.
After a shutdown of this server system have a Red Alert caused by a sensor of the Hardware Status View.
The error that we see is:
1 alert: System Board 0 SEL_FULNESS (out of 142 sensors).
We tried to Reset Sensors, Update information without resolution. Then we try to check detailed information of this alarm and we saw:
++ System Board 0 SEL_FULLNESS ! Alert Reading: 100+
Upper Threshold Crtitical: 80
Anyone know this problem?
Can you help me?
Thanks
Marco
what kind of hardware? Have you confirmed the hardware is in fact not in any sort of degraded state?
You can "acknowledge" the alarm and "reset to green", but that will just make the alarm go away.
Hardware is a Cisco Systems N20-B6620-1 (UCS BLADE) .
If we checki hardware using UCS Java console we don't see any problem and its state is good.
Our problem is that we can't be sure that it's only an aesthetical problem. I can't find any description of the error/problem.
No News, Good News. We have not resolved our problem.
Are there anyone that can help us?
well, if it's only on one system and no other hits on the web for other people seeing this to suggest it's a known false report, I'd take it as an indication of a hardware issue and call Cisco support.
Just because it's still running ok doesn't mean that it isn't suffering from a hardware fault. That's why there's hardware monitoring, because history has shown that systems can be having a problem that doesn't become obvious until the whole system crashes.
Onboard SEL is full, you can empty it from GUI.
SEL is used to store messages/logs from blade. Each blade has one.
Go to Equipment -> Chassis Nr. X -> Server Nr. Y -> Management Logs
and then clear.
wdey is correct. You may need to reset the sensors as well. FYI, clearing the event log from inside the virtual center via "hardware status" view system event log, then "reset event log" will clear this as well.
edit: you may see this coupled with an ipmi error, which is simply reporting that the log is full as well, so both errors are in regards to the same thing.
Nice one wdey