Netra X4200 M2 Servers shipped with illegal memory configuration.

Asset ID:	1-73-1020865.1
Update Date:	2010-08-18
Keywords:

Solution Type FAB (standard) Sure

Solution 1020865.1 : Netra X4200 M2 Servers shipped with illegal memory configuration.

Impact

Some Netra X4200 M2 Servers shipped with illegal/unsupported memory configuration. This configuration may cause intermittent ECC Uncorrectabe Errors (UE) and subsequent server reboots, which may result in unnecessary replacement of parts.

Contributing Factors

This issue impacts some Netra X4200 M2 Servers that have a serial number date code of between 0713xxxx and 0821xxxx.

Affected systems will have the inner DIMM slots pair (A0,B0) populated and the outer slots pair (A1,B1) empty. The Netra X4200 M2 DIMM population rules state that the outer slots must be populated first.

The symptoms are typically not seen until the system is upgraded to firmware 1.4.1, as this firmware incorporates some tuning of the memory timing.

Symptoms

The symptoms of this issue have only been seen on systems with incorrectly positioned Hyundai DIMMs. Systems with other brands of DIMMs may also be incorrectly configured, however the below memory errors are typically not seen.

Not abiding to the Netra X4200 M2 DIMM population rules can potentially cause system instability in many ways. However the most common symptom by far, would be Uncorrectable errors similar to the ones below;

- ipmitool -I lan -H ilom-ip-address -U root sel elist

7000 | 04/01/2009 | 01:44:11 | OEM #0x12 | | Asserted
7100 | OEM record e0 | 00000000041308089f01a048f6
7200 | OEM record e0 | 010000000480af940b00000000
7300 | 04/01/2009 | 01:44:24 | Memory | Uncorrectable ECC | Asserted | CPU 0 DIMM 0
7400 | 04/01/2009 | 01:44:24 | Memory | Memory Device Disabled | Asserted | CPU 0 DIMM 0
7500 | 04/01/2009 | 01:44:27 | System Firmware Progress | Motherboard initialization | Asserted
7600 | 04/01/2009 | 01:44:27 | System Firmware Progress | Video initialization | Asserted
7700 | 04/01/2009 | 01:44:33 | System Firmware Progress | USB resource configuration | Asserted

- ipmitool -I lan -H ilom ip-address -U root sunoem led get all

p0.d0.led        | ON
p0.d1.led        | ON
p0.d2.led        | OFF
p0.d3.led        | OFF
p1.led           | OFF
p1.d0.led        | ON
p1.d1.led        | ON
p1.d2.led        | OFF
p1.d3.led        | OFF

- ipmitool -I lan -H ilom-ip-address -U root fru print

FRU Device Description : p0.fru (ID 6)
Product Manufacturer : ADVANCED MICRO DEVICES Product Name   : DUAL-CORE AMD OPTERON(TM) PROCESSOR 2214 HE
Product Part Number   : 0F41
Product Version       : 03

FRU Device Description : p0.d0.fru (ID 8)
Product Manufacturer : HYUNDAI ELECTRONICS
Product Name          : 2048MB DDR-II 666 (PC2 5300) ADDRESS/COMMAND PARITY/ECC
Product Part Number   : HYMP525P72CP4-Y5 Product Version       : 4141
Product Serial        : 00005069

FRU Device Description : p0.d1.fru (ID 9)
Product Manufacturer : HYUNDAI ELECTRONICS
Product Name          : 2048MB DDR-II 666 (PC2 5300) ADDRESS/COMMAND PARITY/ECC
Product Part Number   : HYMP525P72CP4-Y5 Product Version       : 4141
Product Serial        : 04004075

FRU Device Description : p0.d2.fru (ID 10) < ---- should be populated

FRU Device Description : p0.d3.fru (ID 11) < ---- should be populated

FRU Device Description : p1.fru (ID 7)
Product Manufacturer : ADVANCED MICRO DEVICES Product Name     : DUAL-CORE AMD OPTERON(TM) PROCESSOR 2214 HE
Product Part Number   : 0F41
Product Version       : 03

FRU Device Description : p1.d0.fru (ID 12)
Product Manufacturer : HYUNDAI ELECTRONICS
Product Name          : 2048MB DDR-II 666 (PC2 5300) ADDRESS/COMMAND PARITY/ECC
Product Part Number   : HYMP525P72CP4-Y5 Product Version       : 4141
Product Serial        : 00003074

FRU Device Description : p1.d1.fru (ID 13)
Product Manufacturer : HYUNDAI ELECTRONICS
Product Name          : 2048MB DDR-II 666 (PC2 5300) ADDRESS/COMMAND PARITY/ECC
Product Part Number   : HYMP525P72CP4-Y5 Product Version       : 4141
Product Serial        : 00006066

FRU Device Description : p1.d2.fru (ID 14) < ---- should be populated

FRU Device Description : p1.d3.fru (ID 15)   < ---- should be populated

If you look at the above three outputs you would see that d0 and d1 are populated, which is not a legal configuration.

Root Cause

The instructions for installing DIMMs in the manufacturing process were incorrect until this was highlighted in April of 2008. This would imply that any Netra X4200 M2 system built prior to this could potentially be mis-assembled and could be prone to the issues highlighted by the memory timing changes. Therefore, Sun estimates that the start date for this issue is the date that this product first started shipping.

The manufacturing instructions were corrected and systems with serial numbers later than 0821XXXX are not impacted by this issue. Sun Manufacturing believes this issue was isolated only to the Netra X4200 M2 product. Fully populated Netra X4200 M2 Servers are not impacted by this issue.

Corrective Action

Workaround:

No workaround available - see Resolution section.

Resolution:

Follow the Netra X4200 M2 Service Manual (Chaper 3 Page 72) DIMM population method in order to resolve this issue ;

   http://dlc.sun.com/pdf/820-0063-11/820-0063-11.pdf

The Netra X4200 M2 DIMM population rules state that the outer slots pair must be populated first.

FRU Device Description : p0.d2.fru (ID 10)   < ---- should be populated (A1)
FRU Device Description : p0.d3.fru (ID 11)   < ---- should be populated (B1)
FRU Device Description : p1.d2.fru (ID 14)   < ---- should be populated (A1)
FRU Device Description : p1.d3.fru (ID 15)   < ---- should be populated (B1)

References:

   Escalation ID: 22952533, 22959169
   Radiance ID: 11351344

For information about FAB documents, its release processes, implementation strategies and billing information, go to the following URL:

http://tns.central/fab

For Sun Authorized Service Providers go to:

http://tns.central/SPE/fab/

In addition to the above you may email:

FAB-Manager@sun.com

Internal Contributor/submitter
Sreekumar.Prabhakaran@Sun.COM

Internal Eng Responsible Engineer
Sreekumar.Prabhakaran@Sun.COM Responsible Manager: Takefumi.Hyodo@Sun.COM

Internal Services Knowledge Engineer
Joe.Davis@Sun.COM

Internal Eng Business Unit Group
Systems Group-Netra Systems and Networking Systems Group-x64 (X4100-X4600 (and M2), V20z/V40z/V60z/V65z, Ultra20/40 (and M2) Workstations)

Internal Sun Alert & FAB Admin Info
25-Aug-2009: Completed draft and sent to Extended Review.
28-Aug-2009: All Ext Rvw issues resolved - sending to Publish.
18-Nov-2009: Corrected Product Name to swoRDFish inconsistency.

Attachments

This solution has no attachment