ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

Two brand new DL140 G3's - 3V CMOS Sense problem

SOLVED
Go to solution

Two brand new DL140 G3's - 3V CMOS Sense problem

I have 2 brand new DL140 G3 systems. In either one, the LO100 remote management web page says the system status is critical, and the "monitoring sensors" view has a line marked in red that shows:
3V CMOS Sense - Lower Critical - 1.082v

This sounds like a straightforward hardware problem except that it's happening to two different systems which makes me think something else might be going on. Is there anything I should look at, like flashing the bios, before returning the systems under warranty?

Also, the systems appear to boot OK (at least from CD - no OS installed yet). If this is really a problem, what symptom should I expect to see if the 3V CMOS voltage is low?
2 REPLIES
Ian Grobler
Frequent Advisor

Re: Two brand new DL140 G3's - 3V CMOS Sense problem

J,
I've seen a similar problem on one DL140G3 (out of a batch of 20).
The voltage on the iLO 3V CMOS Sense - Lower Critical is being read as 2.192 Volts.
The server itself is fine and has not rebooted/crashed or had any such problems.

Oddly Insight Manager keeps losing/picking up the iLO during it's polling interval. I tried a continuous ping of the iLO and there's no packet loss or anything like that.
This could be a bug or something more.
iLO firmware details as follows:
IPMI Version: 2.0
Firmware Version: 2.12
Hardware Version: 1.0

DL140G3 is running firmware O08 (18/05/2007).
Hopefully this is just a bug or the iLO may be fauly.. Anyone have any ideas or had similar issues?
Dirk Bergren
Occasional Visitor
Solution

Re: Two brand new DL140 G3's - 3V CMOS Sense problem

I have a system with 57 DL140 G3s, and I've had to replace the CMOS battery on 30 of them. Once the battery is replaced, the low critical condition will go away. You'll need to re-enter the BIOS date and time, and the ILO NIC configuration if you're using static addressing.

This first came to our attention when we were bringing up the servers after shipping the system from our factory to the customer site. 23 of them would not boot and gave us the error message "Kernal Panic not synching:nmi watchdog.Badness in panic at kernal/panic". Aparently, on the servers where the battery was completely dead, the nmi_watchdog got set to a value other than 0. The work around was the add nmi_watchdog=0 to the boot command line.