How to Replace a Sun ZFS Unified Storage Appliance Memory DIMM [VCAP]

Asset ID:	1-71-1408348.1
Update Date:	2018-05-11
Keywords:

Solution Type Technical Instruction Sure

Solution 1408348.1 : How to Replace a Sun ZFS Unified Storage Appliance Memory DIMM [VCAP]

Applies to:

Exalogic Elastic Cloud X4-2 Quarter Rack - Version X4 to X4 [Release X4]
SPARC SuperCluster T4-4 Full Rack - Version All Versions to All Versions [Release All Releases]
SPARC SuperCluster T4-4 - Version All Versions to All Versions [Release All Releases]
Sun ZFS Backup Appliance - Version All Versions and later
Exalogic Elastic Cloud X3-2 Full Rack - Version X3 to X3 [Release X3]
7000 Appliance OS (Fishworks)

Goal

How to Replace a Sun ZFS Unified Storage Appliance Memory DIMM

Solution

DISPATCH INSTRUCTIONS

WHAT SKILLS DOES THE ENGINEER/ADMINISTRATOR NEED:
Knowledge of appliance hardware

TIME ESTIMATE: 90 minutes
TASK COMPLEXITY: 0

PROBLEM OVERVIEW
What: Replace faulty DIMM
Where: DIMM [SR owner to specify], confirm location with BUI:
Maintenance - Hardware - show details - DIMM
Why : DIMM diagnosed faulty

REPLACEMENT INSTRUCTIONS

See References for a list of other useful troubleshooting documents.

PLEASE NOTE: If this is a 7420, ZFS Backup Appliance, ZS3-4 or ZS3-BA please see document Oracle ZFS Storage Appliance: 7420 and ZS3-4 DIMM mis-identification on fault (when running ILOM/BIOS 3.0.16 and higher) (Doc ID 2286134.1) before continuing with this procedure.

WHAT STATE SHOULD THE SYSTEM BE IN TO BE READY TO PERFORM THE RESOLUTION ACTIVITY:
* Shut down appliance
* Disconnect power cords.

Note: For Cluster systems the resources must all be on the other node.

NOTE 2: For Clustered Systems

It is MANDATORY NOW THAT YOU UNPLUG THE CLUSTRON CABLES (UART0 and UART1 ) and take notice that these cables are crossover connected.

THEY WILL NEED TO BE REPLUGGED IN THE SAME WAY AFTER THE HW REPLACEMENT ACTION HAS BEEN COMPLETED.

WHAT ACTION DOES THE ENGINEER/ADMINISTRATOR NEED TO TAKE:
1 Take antistatic precautions
2 Remove system from rack
3 Remove top cover - press release button, slide slightly to rear before removing
* If servicing a 7410, lift the air duct to replace DIMMs on the mezzanine board, or remove the mezzanine tray to replace DIMMs on the motherboard

4 Remove DIMM
* To help identify faulty DIMM, press Fault Remind button on motherboard, a corresponding amber LED is lit
* If servicing a 7420, lift memory riser that hosts faulted DIMM straight up to remove it from motherboard
* Rotate both DIMM slot ejectors outward to partially eject the DIMM
* Lift DIMM straight up to remove

NOTE for ZS5-2: The memory replacement in ZS5-2 requires to lift the drive cage and also to unplug the SATA cables.

See : https://docs.oracle.com/cd/E79446_01/html/E79459/gpxkg.html#scrolltoc

See also Doc ID 2385369.1 (Oracle ZFS Storage Appliance: How to replace an Oracle ZFS Storage ZS5-2 DIMM)

5 Install DIMM
* DIMMs must be installed in matched pairs in 7x10 and 7x20 systems: ( 7320, does have a single DIMM not in pair)
7x10 paired slots are 0/1, 2/3, 4/5, and 6/7. Memory sockets are colored black or white to indicate which slots are paired. Each pair of DIMMs must be identical (same manufacturer/size/speed)
7420 paired slots are D0/D4 (blue slots), D2/D6 (white slots), D1/D5 (black slots) and D3/D7 (green slots). DIMMs populated in same-colored slots must have same part number

ZS3-4: The ZS3-4 controller supports 16GB DDR3 DIMMs installed on all eight risers,accommodating 1TB of memory.

On every memory riser, slots D0, D2, D4, and D6 must be populated; optionally, slots D1, D3, D5, and D7 may be populated as a group on all installed memory risers. All DIMMs in the system must be identical.

ZS4-4: The ZS4-4 controller has four Intel Xeon E7-8895 v2 15-core 2.8 GHz CPUs and eight memory riser cards . Each memory riser card contains twelve DIMM slots.

ZS5-2: The controller supports only fully populated DIMM's, 16 GB for 384 GB or 32 GB for 768 GB. Do not mix DIMM capacities; use either all 16 GB or 32 GB DIMMs.

7320: Populate the DIMM slots for each memory channel that are the farthest from the CPU first.

For example, populate D8/D5/D2 first; then D7/D4/D1 second; and finally, D6/D3/D0

* Ensure that ejector tabs are fully open
* Align notch in the DIMM with the key in connector
* Push DIMM into connector until ejector tabs lock DIMM in place
* On 7420 push memory riser module back into memory riser slot until it locks in place

6 Return system to operational state
* Place top cover on chassis so that it overhangs rear of the unit slightly. Slide forward until it seats
* Close fan cover and engage fan cover latches. Cover must be completely closed for the system to power on
* Reinstall appliance back in original rack location
* Reconnect all cables
* Verify standby power is on - Power/OK LED flashes on front panel about 2 minutes after power cords are plugged in
* Use a pen to press the recessed Power button on the front panel. The Power/OK LED will stay solidly lit. SSH to the Service Processor and check appliance power state
show /SYS power_state

It would be in off state, before starting power on,check and make sure any faults in ILOM.

Please make sure to see there no faults reported from the ILOM, on the recently replaced DIMM.

If there any please clear it.

Steps:

login to the ILOM CLI and launch a Fault Management Shell session
-> start /SP/faultmgmt/shell
Are you sure you want to start /SP/faultmgmt/shell (y/n)? y

faultmgmtsp>

use the "fmadm faulty -a" command to list out all active faulty components
faultmgmtsp> fmadm faulty -a

If the DIMM that was just replaced is still listed as a fault then indicate that the DIMM has been replaced by using the command "fmadm replaced <fru|cru|uuid>" for example:
faultmgmtsp> fmadm replaced /SYS/MB/P0/D2

confirm that the faults are cleared and then exit out of the fault management shell
faultmgmtsp> fmadm faulty -a
No faults found
faultmgmtsp> exit
->

Now start the system

->start /SYS

OBTAIN CUSTOMER ACCEPTANCE
Confirm appliance is in normal state. Check both BUI and client data access

WHAT ACTION DOES THE ENGINEER/ADMINISTRATOR NEED TO TAKE TO RETURN THE SYSTEM TO AN OPERATIONAL STATE:
Confirm there are no faults reported in appliance BUI
Maintenance - Problems ("Mark Repaired" any prior related DIMM faults)
Should an error return for the same DIMM, reseat the DIMM and monitor the appliance for further errors

PARTS NOTE:
Check System Handbook for correct part number

Memory DIMMs:
7110 : F540-7599 2 x 2GB DIMM
F540-7829 2 x 2GB Single Rank DIMM

7210 :    F541-1313 2 × 2GB DIMM
               F541-1304 2 × 4GB DIMM
               F541-3360 2 × 2GB DIMM

7310 : F540-7600 2 × 4GB DIMM

7410 :    F540-7599 2 × 2GB DIMM
               F540-7829 2 × 2GB Single Rank DIMM
               F540-7600 2 × 4GB DIMM
               F541-3419 2 × 8GB Dual Rank DIMM

7120 :    371-4966 8GB DDR3L-1333/DDR3L-1600 DIMM, 1.35V
               371-4898 4GB DDR3-1333 DIMM, 1.5V
             371-4965 4GB DDR3-1333 DIMM, 1.35V

7320 :    371-4966 8GB DDR3L-1333/DDR3L-1600 DIMM, 1.35V
               371-4898 4GB DDR3-1333 DIMM, 1.5V
               371-4965 4GB DDR3-1333 DIMM, 1.35V

7420 :    X8504A 8GB (2 × 4GB DDR3-1066 DIMMs)
               X8505A 16GB (2 × 8GB DDR3-1066 DIMMs)
               542-0434 2 x 16GB 1.35V DDR3-1066 DIMMs (4-rank)

We have now 2-Rank DIMMs for 7420. Newer 7420 M2 has this pre installed.

                7011549 2 × 4GB DDR3-1333 DIMMs (2-rank)
                7011550 2 × 8GB DDR3-1333 DIMMs (2-rank)
                7059236 2 × 16GB DDR3-1066 DIMMs (2-rank)

4-rank and 2-rank DIMM's shouldn't be intermixed in the same riser.

ZS3-2 : 7053841 32GB DDR3-1066/DDR3-1333 DIMM, 1.35V
7041588 16GB DDR3-1600 DIMM, 1.35V

ZS3-4 : 7059236 32GB Memory Assembly (2 × 16GB DDR3-1066 DIMMs) (2-rank)

ZS4-4 : 7076173 16GB 1.35V DDR3L-1600 DIMM
7073781 32GB 1.35V DDR3L-1600 Load Reduced DIMM

ZS5-2 : 7097801 16GB DDR4-2133 Registered DIMM
7097802 32GB DDR4-2133 Registered DIMM

ZS5-4 : 7095790 16GB DDR3L-1600 1.35V Registered DIMM

To report errors or request improvements on this procedure, please add a comment

Save

References

<NOTE:1277396.1> - Sun x86 Platforms: Matrix of Recognized Device Paths
<NOTE:1324432.1> - Oracle ZFS Storage Appliance: Memory Troubleshooting: The system may have been reset
<NOTE:1556452.1> - How to determine which 16GB 4-rank or 2-rank DIMM to use in X2-8 (formerly X4800M2) CMODs, or X2-4 (formerly X4470M2) memory risers.
<NOTE:1416406.1> - Sun ZFS Storage Appliances Troubleshooting Resource Center
<NOTE:1432269.2> - Information Center: Sun ZFS Storage Appliances
<NOTE:1400578.1> - Sun Storage 7000 Unified Storage System: How to collect SP/ILOM diagnostic data
<NOTE:1402545.1> - Sun Storage 7000 Unified Storage System: How to Troubleshoot Cluster Problems

Attachments

This solution has no attachment