Failover Settings
Using this feature, you can configure a backup server to resume operations in case the system encounters a failover.
Note: Failover must be configured before creating user accounts to avoid the risk of losing data.
Pre-requisites for failover configuration
The hardware and software of the primary and secondary servers must be identical in all aspects. For example, they must have the same type of GPU, same GPU Slot IDs, and the same Userful software version running on them.
The primary and secondary servers should be initialized with the same admin password.
Create a failover group
Go to Admin Center > System Configuration > Failover Settings.
Enter the Cluster Name.
Select the Server Type. The primary server is the operational server that stores all files and manages processing. The secondary server is the backup server.
Enter the IP address and the system ID of the other configured server in the failover server cluster.
The Failover Status shows the current status of this failover cluster.
Click Save Changes.
Please ensure that failover is configured properly before creating user accounts. After configuring failover, you can start creating users.
Types of Failover and Expected Behaviors
Failover type | Happens when: | Expected behavior |
---|---|---|
Manual failover
Failover happens on the active server (primary or secondary whichever is active at the time of triggering failover).
| A user navigates to System Configuration > Failover Settings and clicks Manual failover. | About two to three minutes after failover is triggered, the uClients along with the content on displays restart. |
Power-related failover on the primary server |
|
Note: The content does not move to the primary server when it recovers after power loss. The content remains on the secondary server.
|
Network loss on the primary server while it is the active server |
|
|
Network loss on secondary server while it is on standby |
|
|
Note:
In all of the failover events described above, content restart is not expected to happen more than once.
As long as there are no failover events, content restart is not expected to happen.
Last updated