How to Replace an Exadata X4-8, X5-8, X6-8 Compute Node Memory DIMM

Asset ID:	1-71-2162087.1
Update Date:	2018-05-10
Keywords:

Solution Type Technical Instruction Sure

Solution 2162087.1 : How to Replace an Exadata X4-8, X5-8, X6-8 Compute Node Memory DIMM

Applies to:

Exadata X5-8 Hardware - Version All Versions to All Versions [Release All Releases]
Exadata X4-8 Hardware - Version All Versions to All Versions [Release All Releases]
Exadata X6-8 Hardware - Version All Versions to All Versions [Release All Releases]
Information in this document applies to any platform.

Goal

How to Replace an Exadata X4-8, X5-8, X6-8 Compute Node Memory DIMM.

Solution

DISPATCH INSTRUCTIONS

WHAT SKILLS DOES THE FIELD ENGINEER/ADMINISTRATOR NEED?:
Exadata X4-8/X5-8/X6-8 Training

TASK COMPLEXITY: 3

TIME ESTIMATE: 60 minutes

FIELD ENGINEER/ADMINISTRATOR INSTRUCTIONS:

PROBLEM OVERVIEW: An Exadata X4-8/X5-8/X6-8 Compute Node Memory DIMM needs replacement

WHAT STATE SHOULD THE SYSTEM BE IN TO BE READY TO PERFORM THE RESOLUTION ACTIVITY? :

System should be powered down. Please follow shutdown instructions in DOC ID: 1982342.1

WHAT ACTION DOES THE FIELD ENGINEER/ADMINISTRATOR NEED TO TAKE?:

Reference Doc:

Sun Server X4-8 Service Manual
http://docs.oracle.com/cd/E40591_01/html/E40317/index.html

Oracle Server X5-8 Service Manual
http://docs.oracle.com/cd/E56301_01/html/E56311/index.html

DIMM Overview
http://docs.oracle.com/cd/E40591_01/html/E40317/gofsr.html#XFESMgogco

Memory Installation Notes

1. Populate the slots with black slots and black levers first: D0, D3, D6, D9, D12, D15, D18, D21. The slot population for the minimum DIMM configuration is D0/D6/D12/D18 per CMOD
2. Populate the slots with black slots and white levers second: D1, D4, D7, D10, D13, D16, D19, D22.
3. Populate the white slot, white lever slots last: D2, D5, D8, D11, D14, D17, D20, D23

How to Remove a DIMM

1. Prepare the server for service, shutdown OS and power off server.

2. Remove the 4x fan modules.

3. Remove the fan frame.

4. Remove the CMOD from the server.

5. Set the module on a flat antistatic surface with ample surrounding space and light.

6. Remove the CMOD cover.

7. Press the "DIMM FAULT REMIND BUTTON". This should turn on the slot LED for the faulty DIMM.

8. To unlock the DIMM, simultaneously rotate the two release levers outward fully away from the DIMM.
This action unlocks and ejects the DIMM from the DIMM slot.

9. Remove the DIMM from the CMOD.

How to Install a DIMM

1. Ensure that the DIMM slot locking levers are in the fully open position.

2. Align the DIMM within the slot.
The DIMM is notched to accommodate the key (protrusion) in the slot. The key ensures correct DIMM installation. The DIMM can only be correctly installed one way.

3. To install the DIMM, simultaneously press down on both ends of the DIMM to push it into the slot.
This action causes the DIMM locking levers to lift and lock into place on the DIMM.

4. Verify that the DIMM is properly installed and locked.
When properly locked in place, the DIMM cannot be removed.

5. Install the CMOD cover.

6. Install the CMOD into the server.

7. Install the fan fame.

8. Install the 4x fan modules.

9. After removing a CMOD the DPCCs associated with the CMOD removed/reinserted must also be reseated to ensure they have not become slightly unseated.

OBTAIN CUSTOMER ACCEPTANCE
WHAT ACTION DOES THE FIELD ENGINEER/ADMINISTRATOR NEED TO TAKE TO RETURN THE SYSTEM TO AN OPERATIONAL STATE?:

Prepare server for operation, power on server and start OS. Please follow start up instructions in DOC ID: 1982342.1

How to verify the DIMM is working properly

Log in to the ILOM CLI.

Enter the following command to check status is normal status:

-> show /System/Memory/DIMMs/DIMM_x
note: the "x" represents the DIMM number of the DIMM replaced

Example:

-> show /System/Memory/DIMMs/DIMM_0

/System/Memory/DIMMs/DIMM_0
    Targets:

    Properties:
        health = OK
        health_details = -
        part_number = 001-0003-01,M393B2G70DB0-YK0
        serial_number = 00CE011412225F9C48
        location = P0/D0 (CPU 0 DIMM 0)
        manufacturer = Samsung
        memory_size = 16 GB

Use the fault management shell to clear the fault for a specific component:

For example:

faultmgmtsp> fmadm acquit /SYS/MB/P0/D0

or you can use the UUID

faultmgmtsp> fmadm acquit UUID

Check if any error output from event log

Example

-> show /SP/logs/event/list

Check if any faulted parts exist.

-> start /SP/faultmgmt/shell
Are you sure you want to start /SP/faultmgmt/shell (y/n)? y

Example

faultmgmtsp> fmadm faulty
No faults found

Check FMA and OS information to verify that no errors existed

Example

# fmadm faulty -a
STATE RESOURCE / UUID
-------- ----------------------------------------------------------------------

# prtdiag -v
System Configuration: Sun Server X4-8
BIOS Configuration: American Megatrends Inc. 29011300 08/26/2010
BMC Configuration: IPMI 2.0 (KCS: Keyboard Controller Style)

==== Processor Sockets ====================================

Version Location Tag
-------------------------------- --------------------------
Intel(R) Xeon(R) CPU E7-8895 v2 @ 2.80GHz CPU 1
Intel(R) Xeon(R) CPU E7-8895 v2 @ 2.80GHz CPU 2
Intel(R) Xeon(R) CPU E7-8895 v2 @ 2.80GHz CPU 3
Intel(R) Xeon(R) CPU E7-8895 v2 @ 2.80GHz CPU 4

==== Memory Device Sockets ================================

Type Status Set Device Locator Bank Locator
------- ------ --- ------------------- --------------------
SDRAM in use 0 BL0/P0 DIMM0<------here
SGRAM empty 0 BL0/P0 DIMM1
SDRAM in use 0 BL0/P0 DIMM2
SGRAM empty 0 BL0/P0 DIMM3
SDRAM in use 0 BL0/P0 DIMM4
.
.

Check /var/adm/messages file to verify that no errors existed

Example

bash-3.00# pwd
/var/adm

# grep -i warning messages
# grep -i error messages
# grep -i fail messages

PARTS NOTE:

REFERENCE INFORMATION:

Sun Server X4-8 Service Manual
http://docs.oracle.com/cd/E40591_01/html/E40317/index.html

Oracle Server X5-8 Service Manual
http://docs.oracle.com/cd/E56301_01/html/E56311/index.html

DIMM Overview
http://docs.oracle.com/cd/E40591_01/html/E40317/gofsr.html#XFESMgogco

How to Shutdown and Startup Exadata X5 compute nodes and storage cells when performing hardware maintenance (includes Supercluster X5 storage cells) (Doc ID 1982342.1)

References

<NOTE:1381773.1> - How to clear FMA logs on the ILOM or Solaris on x86 platforms

Attachments

This solution has no attachment