Upgrading to ESH or ESH2 modules

DS14 (only available at 1-Gb loop speed)

LRC to ESH or ESH2 hot-upgrade is supported.

ESH or ESH2 to LRC hot-downgrade is supported.

DS14mk2 FC set at
1-Gb loop speed

LRC to ESH or ESH2 hot-upgrade is supported.

ESH or ESH2 to LRC hot-downgrade is supported.

DS14mk2 FC set at
2-Gb loop speed

• Not applicable
• ESH to ESH2 is supported. SeeHot-swapping a module.

• ESH or ESH2 to LRC hot-downgrade is not supported.
• ESH2 to ESH hot-downgrade is supported.

Determine which loop you will upgrade first, and determine whether any disks in the cluster are single-pathed through that loop.

For more information, see How to determine path status for your cluster.

If any disks use this loop for their only path to a controller, complete the following steps:

1. On one node, which we will call Node A, enter the following command:

   cf takeover
   

2. Wait for takeover to be complete and make sure that the partner node, or Node B, reboots and is waiting for giveback.

Result

You can now replace all modules on all loops attached to Node B.

Note the terminate switch position of the LRCs you are replacing in this loop. If the new modules have terminate switches, they should be set to the same setting as the LRCs they are replacing.

Note
The ESH2 is self-terminating and does not have a terminate switch.

While grounded, unplug the cabling from the dormant loop and note the cable locations.

Remove the LRC modules from the disk shelves that do not have a cable attached.

Insert all of the new modules.

Attention
After inserting all modules, wait 30 seconds before proceeding to the next step.

Recable the disk shelves to their original locations.

Check the operation of the new modules by entering the following command from the console of the node that is still running:

environ shelf

Result

The node reports the status of the modified shelves as good.

If there is more than one loop attached to the node that has been taken over, repeat Step 3 through Step 8 for all other loops on that node.

If you performed a cluster takeover previously, complete the following steps:

1. At the console of the takeover node, return control of Node B's disk shelves by entering the following command:
   
   cf giveback
2. Wait for the giveback to complete before proceeding to the next step.

Check the operation of the new modules by entering the following command on the console of both nodes:

environ shelf

Result

Each node reports the status of the modified shelves as good.

Check the next loop you will upgrade to determine whether any disks in the cluster are single-pathed through that loop.

For more information, see How to determine path status for your cluster

If any disks use this loop for their only path to a controller, complete the following steps:

1. On Node B, enter the following command:

   cf takeover
   

2. Wait for takeover to be complete and make sure that the partner node, or Node A, reboots and is waiting for giveback.

Result

You can now replace all modules on all loops attached to Node A.

Repeat Step 3 through Step 11, with Node B as the takeover node.

Test the upgraded modules, as described in Testing the modules.

Test the configuration, as described in Stage 6: Checking and testing cluster configuration.