Table of Contents
Advertisement
Before you begin
• If you have a cluster with more than two controllers, check the health and Epsilon from advanced mode:
cluster show -epsilon*
• If the cluster is not in quorum or a controller that is not the degraded controller shows
and health, correct the issue before proceeding to the next step.
• If Epsilon resides in the degraded controller:
1. Remove Epsilon from the degraded controller:
2. Assign Epsilon to a healthy controller in the cluster:
Step 1.
If the degraded controller is part of an HA pair, disable automatic giveback from the console of the
healthy controller using the following command:
false
Step 2.
Take the degraded controller to the LOADER prompt by typing
If the degraded controller is displaying...
The LOADER prompt
Waiting for giveback...
System prompt or password prompt
Step 3.
Turn off the power supplies, and then unplug the degraded controller's power cords from the
power source.
Replacing the controller module hardware
To replace the controller module hardware, you must remove the degraded controller, move FRU
components to the replacement controller module, install the replacement controller module in the chassis,
and then boot the system to Maintenance mode.
ThinkSystem DM3000x and DM5000x Hardware Installation and Maintenance Guide
48
cluster modify -node degraded_node -epsilon false
cluster modify -node healthy_node -epsilon true
storage failover modify -node local -auto-giveback
.
halt
Then...
Go to the next step.
Press Ctrl-C, and then respond
prompted.
Take over or halt the degraded controller:
• For an HA pair, take over the degraded
controller from the healthy controller:
failover takeover -ofnode impaired_node_name
When the degraded controller shows
Waiting for giveback..., press Ctrl-C, and
then respond
y
for eligibility
false
when
y
storage
.
Advertisement
Table of Contents
