With the Return Material Authorization (RMA) process, you can replace a faulty device
with a new device that has the same configuration.
The high-level process is as follows. For specific steps and commands, see Replace a Faulty Device.
Verifying prerequisites.
Periodic configuration backup
must be enabled on all devices that may need RMA. This prerequisite ensures
that you have the latest configuration file to be used for recovery.
Maintenance mode must be
enabled upon reboot on all devices.
Removing the faulty device and replacing it
with the new device. Ports on this device must be administratively up and
online.
Configuring the new device with the same management IP address and credentials as the
old device.
Starting the RMA process from the command line
or with the REST API.
Note
As a best
practice, run the efa
inventory rma execute command with the configuration backup ID so
that the configuration is properly restored. If you run the command without the
backup ID, you will need to manually update the configuration on the new
device.
During the RMA process, the following actions occur:
The device boots up in maintenance mode.
EFA updates the device ID for the connection details in the
database.
Maintenance mode is initiated if the device is not already in
maintenance mode.
EFA replays the backed-up configuration specified
by the config-backup-id parameter of the efa
inventory rma execute command.
EFA begins the drift reconcile process, which
involves device discovery, device update, and fabric and tenant
reconciliation. For more information, see Drift
and Reconcile.
Note
If the RMA command fails during this
stage, you can manually run the drift reconcile process from the
CLI. If the RMA process fails for any other reason, restart the RMA
process.
When drift reconcile is complete, the device is taken out of maintenance
mode.
During the RMA process, EFA health checks are deactivated and RASlog
does not trigger drift reconcile.
Installing the HTTPS or OAuth2 certificate on the new device.