The system controller replace start command is used to initiate a controller-replacement operation. The update is preceded by a validation of the HA pair to ensure that any issues that might affect the update are identified.

There are predefined points in the update when the update can be paused (either requested by the user or by the operation in case of an error or for manual intervention).

Parameters

[-nodes <nodename>,…] - Nodes for Controller Replacement: Specifies the nodes that are to be replaced.
[-simulate <true>] - Simulate Controller Replacement: Dry run the operation. Checks for all validations.
[-skip-metrocluster-check {true|false}] - Skip Metrocluster Check before Replacement: Skips the DR mirror offline check when performing Metrocluster validation. In 4-node Metrocluster configuration, if controller replacement is already complete one one site, then the partner site should replace its controllers by setting this parameter to true . The default value is false.
[-nso {true|false}] - Select NSO Procedure: Follow the negotated switchover switchback based controller replacement procedure for 4 node MCC FC system. To select ARL based procedure for 4 node MCC FC , this parameter needs to be set to false .This parameter is only honored in MCC configuration. Hence, setting or unsetting of this parameter has no effect in HA configuration. The default value is true for MCC FC.

Examples

The following example shows the replacement operation:

cluster1::> system controller replace start -nodes node1,node2 -simulate true

Warning: 1. Current version of node is 9.4.0
Before starting controller replacement, please ensure that the new controllers are in the version 9.4.0
2. Verify that NVMEM or NVRAM batteries of the new nodes are charged, and charge them if they are not. You need to physically check the new nodes to see if
the NVMEM or NVRAM batteries are charged. You can check the battery status either by connecting to a console or using SSH, logging into the Service Processor
(SP) for your system, and use the system sensors to see if the battery has a sufficient charge.
Attention: Do not try to clear the NVRAM contents. If there is a need to clear the contents of NVRAM, contact Fujitsu support personnel.
3. If you are replacing the controllers with an used one, please ensure to run wipeconfig before controller replacement
4. Current model name is FAS8040
Before starting the operation, please ensure that the new controller model is supported for controller replacement.

Do you want to continue? {y|n}: y
Controller replacement: Prechecks in progress...............
Controller replacement has been paused for user intervention.
Please collect the following info from the current node:
vserver services name-service dns show
network interface show -curr-node node -role cluster,intercluster,node-mgmt,cluster-mgmt
network port show -node node -type physical
service-processor show -node * -instance
network fcp adapter show -node node
network port ifgrp show
network port vlan show
system node show -instance -node node
run -node node sysconfig
storage aggregate show -node node
volume show -node node
network interface failover-groups show
storage array config show -switch switchname
system license show -owner node
storage encryption disk show