RX5670, howto identify A DIMM that needs replacement.

[bennethos]

Hi all,

I have an rx5670 and have the following logs in the SM console :

Log Entry 106: 10 Aug 2009 11:14:47
Alert Level 2: Informational
Keyword: MEM_CORR_ERR
Correctable (javascript:postMessageSubmit('submit');single-bit) ECC error in DIMM
Logged by: System Firmware 0
Data: Location - Memory (SIMM or DIMM): DIMM Slot 0xC, Extender 0

Slot 0xC is my problem, I have lots of dimms in that machine, too much to trial and error ;-). I was also thinking of changing the whole bank of dimms...

thx for helping me out !

[Torsten.]

I think there was a hex code next to the message you posted. However, the DIMM is on extender 0, the slot should be labeled 0c.

Is this system running hp-ux?

Try this:

# echo "selclass qualifier memory;info;wait;infolog" | /usr/sbin/cstm

[Stefan Stechemesser]

Hi,

take a look at the service manual for the Dimm locations (you message is regarding Dimm 0xC on Extender 0)

http://docs.hp.com/en/rx5670_OpsMaint/ch06s20.html

But: normally it is not nessecary to replace a dimm due to correctable memory errors unless you would see a very high amount of them on the same Dimm.

If the Dimm has a severe problem, it will be deconfigured during memory selftest.

best regards

Stefan

[bennethos]

thx for the help so far.

That's what I thought as well, but there are lots of ECC errors and there's a windows running on the machine (not my choice ! :p ), the machine is also showing excessive CPU CACHE checks for all cpu's in event logs of windows.He's constantly checking it seems...

I think after a dimm gives an ecc error 4 messages are spawned in the log concerning checking the cache of all cpu's