Question on failback-delay for Artemis 2.19.1 shared store config

[Gunawan, Rahman (GSFC-703.H)[BUSINESS INTEGRA, INC.]]

I'm trying share store configuration based on this doc 
https://activemq.apache.org/components/artemis/documentation/1.0.0/ha.html.  
Failback-delay is configured to be 300000 (300 seconds) hoping that when 
primary goes down, the backup takes over and when the primary is back alive, it 
waits for 300 seconds before the primary takes over.  However; the primary took 
over immediately when I started it.  Was the config wrong?
Is there a way to let the backup server as live server when primary goes down 
and stay that way although the primary recovers from crash (primary will be 
announced as backup)?

Below is the configuration for primary and secondary:
The primary is configured below:
      <ha-policy>
         <shared-store>
            <master>
               <failover-on-shutdown>true</failover-on-shutdown>
               <failback-delay>300000</failback-delay>
            </master>
         </shared-store>
      </ha-policy>

The backup is configured:
      <ha-policy>
         <shared-store>
            <slave>
               <failover-on-shutdown>true</failover-on-shutdown>
               <allow-failback>true</allow-failback>
               <failback-delay>300000</failback-delay>
            </slave>
         </shared-store>
      </ha-policy>

Thanks

Regards,
Rahman