How do I move a CLARiiON array from one cabinet or rack to another?


CLARiiON CX Series

Disk Library DL Series



How do I move a CLARiiON array from one cabinet or rack to another?

Requirements for moving a CLARiiON array from one cabinet to another

How to disassemble/reassemble a CLARiiON array



1.       Preparations/Warnings/Assumptions:

a. Preparation: If Possible, upgrade code to latest revision, verify utility partition and recover image are installed.

b. Warning:  Never move an array that has any fault.  Always correct all faults prior to moving array.

c. Assumption:  All hosts have been disconnected and cables properly labeled.

d. Assumption:  Front-end ports are disconnected from switch/host and cables are properly labeled.


2.       Check overall health.

a.       Verify that there are no faults or unowned LUNs (except hot spares) on the array from Navisphere       GUI and disable Read and Write Cache.


3.       Document basic array information using Navisphere CLI.

a.       Execute navicli h <SPA_IP_address> getagent

b.       Execute navicli h <SPB_IP_address> getagent

c.        Verify both commands returned the same information noted below. If this information is NOT the       same, the IP addresses supplied were incorrect or the SPs are running two different versions of PROM/FLARE code and this situation MUST be corrected first before proceeding.


d.       Serial No : _________________________

e.       Model : _________________________

f.        Revision : _________________________

g.       PROM Rev : _________________________


h.       SP Identifier: Should report A  for SPA_IP_address and B for SPB_IP_address


4.       Document current array configuration using Navisphere CLI.

a.       Gather SPcollect files by referring to solution 5253.

b.       Save these files somewhere safe and easily accessible.


5.       Document disk location by serial number using Navisphere CLI.

a.       Execute navicli h <SPA_IP_address> getdisk -serial >getdisk_SPB.txt

b.       Execute navicli h <SPB_IP_address> getdisk -serial >getdisk_SPB.txt

c.        Save this file somewhere safe and easily accessible.


6.       Document enclosure and LCC Location by serial number using Navisphere CLI.

a.       Execute navicli h SPA_IP_address> getresume -messner >getresume_SPA.txt

b.       Execute navicli h <SPB_IP_address> getresume -messner >getresume_SPB.txt

c. Save this file somewhere safe and easily accessible.


7.       Document overall general array health.

a.       Execute navicli h SPA_IP_address> getcrus -messner >getcrus_SPA.txt

b.       Execute navicli h <SPB_IP_address> getcrus -messner >getcrus_SPB.txt

c.        Edit each of these .txt files and verify the words FAULT,  EMPTY, NOT PRESENT, REMOVED, UNKNOWN," or INVALID do not appear ANYWHERE except for where expected. As an example, a CX300 array with a single standby power supply (SPS) might report that one of them is EMPTY since there is only one. This would be OK.


8.       Physically label every enclosure.

a.       Verify each enclosure address that is present by using the following command:

navicli -h <SPx_IP_address> flashleds e [0-7] b [0-3] on/off

b.        Do not forget to TURN the LEDS OFF!!!

c.       It is best to execute this command to SP-A, turn the enclosures on then off, and then repeat for SP-B. This is to test that the existing backend cabling is correct. If there is a discrepancy, STOP and correct or call in for assistance if needed.

e.       Label the enclosure :


ADDRESS 0_1 (Formed from Bus and Enclosure B_E)


9.       Review both SP event logs and verify that no major issues are occurring or have occurred in previous 24 hours.

10.    Procedure Status Check: All enclosures labeled and current configuration documented via Navisphere CLI, array is healthy. Verify NO FAULTS exist on the array.

11.    At this point you can power off the array.

12.    Physically move equipment.

a.       Refer to related hardware reference manuals to familiarize your self with how to remove/install rail kits or rack mount kits.


13.    PRIOR to powering equipment ON.

a.       If disks were removed, use the information in the getdisk_SPX.txt file (Step 5) and compare with the enclosure label that you applied (Step 8) to ensure that every disk is placed in their proper slot.

b.       Refer to the specific array model hardware Setup Guide and verify that the back-end enclosure cabling is in proper order.

c.        An example for the CX700 Reference manual is :CX700 2-Gigabit Fibre Channel Storage Processor Enclosure (SPE) Setup Guide  (P/N 300-001-274-Axx)


14.    Warnings:

a.  Enclosure 0_0 contains the BOOT OS and CONFIGURATION for the array. If this enclosure is not present the array will not come up at all.

b.  DO NOT interchange Enclosure 0_0 from one physical array to another.

c.  Keep all original enclosures for a specific array together until array is powered up and healthy.

d.  DO NOT alter the hardware configuration (add/remove enclosures or disks) until AFTER the array powers up and is healthy.

e.  Any Disks that contained DEFINED raid groups must have the same exact address as they had prior to powering the array off. The array remembers what disks were in use , where and how they were used, prior to being powered down and will need to see these disks at  the same address on power-up. Otherwise data unavailability/data loss can occur.


15.    Verify hardware cabling, and enclosure addresses, disks, and power on equipment.

16.    Verify array is healthy and then enable Read and Write Cache.


For more information, Refer to EMC Solution Article 7835