DL980 Memory error

[Mastest]

Hi friends

I have one server Dl980 G7 and 2Tb memory installed. I install oracle linux on this and when i create table space on this server in oracle database the server send a lot of error .

MCA: MEMORY CONTROLLER GEN_CHANNELunspecified_ERR
Transaction: Generic undefined request
STATUS 900000400009008f MCGSTATUS 0
MCGCAP 1000c16 APICID e0 SOCKETID 7
CPUID Vendor Intel Family 6 Model 46
Hardware event. This is not a software error.
MCE 0
CPU 56 BANK 9
TIME 1341330227 Tue Jul  3 20:13:47 2012
MCG status:
MCi status:
Error overflow
Corrected error
Error enabled
MCA: MEMORY CONTROLLER GEN_CHANNELunspecified_ERR
Transaction: Generic undefined request
STATUS d0000080000a008f MCGSTATUS 0
MCGCAP 1000c16 APICID e0 SOCKETID 7
CPUID Vendor Intel Family 6 Model 46
Hardware event. This is not a software error.
MCE 0
CPU 0 BANK 9
TIME 1341330362 Tue Jul  3 20:16:02 2012
MCG status:
MCi status:
Corrected error
Error enabled
MCA: MEMORY CONTROLLER GEN_CHANNELunspecified_ERR
Transaction: Generic undefined request
STATUS 900000400009008f MCGSTATUS 0
MCGCAP 1000c16 APICID 0 SOCKETID 0
CPUID Vendor Intel Family 6 Model 46
Hardware event. This is not a software error.
MCE 0
CPU 0 BANK 9

In hardware diagnostic board don't send any error .

Please help me

[Suman_1978]

Hi,

Its complaining about Bank 9, tell us what type of processor this server has and what memory modules it has, I mean the combination of DIMM in slots.

"CPUID Vendor Intel Family 6 Model 46", by this I am guessing its a Intel X7560 2.27 GHz/8-core CPU, correct me if wrong.

Have you performed Insight Diagnostics from the SmartStart CD by booting into SmartStart CD, if not then try it.

Thanks

Suman AN

[Mastest]

The CPU model is X6550 and memory model is 16Gb 8500 PC3.

When i running system diagnostic, i have free error report. I do this work any times and every time the system error free.

Regards

[luc simard]

Hello Mastest, these are CORRECTABLE MCEs that occur when the CPU transitions from different c-States (has to do with power saving features of the CPU). As the system transitions from lower power state, it needs to "wake up" the sleeping resources and this is why you see the MCE errors. They tend to fill the MCE logs but can be disregraded as there is no action to be taken as a result of those errors. You will not see those in the IML since they are correctable and no action needs to be taken. Diags will always come clean with no issues to report.

You can refer to this customer advisory for the DL580 (same issues as described for the DL980):

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c03282091

and this one:

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c03356780

As the Advisory describes, you can disable the C-State features within the RBSU. This will prevent the server from saving energy while idle, but will stop logging the correctable MCEs.

Hope this Helps

Luc