Cant fail over primary to standby 11.5 portal

JeffreyGrussing-GRE

Our development HA portal wont fail over properly from primary to stand by. here is the issue I am having.

Portal 1 is primary Portal 2 is standby is i stop the portal service on Portal 1 to fail it to Portal 2 it takes the portal site down and it doesn't come back until I start Portal 1 back up then the Portal comes back and in Portal Admin it shows Portal 2 as primary and Portal 1 is standby just like its suppose to and both check healthy. But with Portal 2 as Primary it requires Portal 1s service to be up in order for the site to be up. I have done the unregister reregister and that didn't fix the problem. anyone know what I need to do. Our production enterprise site works just fine but our dev site has this issue.

Thanks Jeff

TimoT

Hi @JeffreyGrussing-GRE 

When Portal 1 is down, are you still able to access the portal home page directly on the Portal 2 machine via port 7443 URL (https://portal-machine.domain.com:7443/arcgis/home)?

Overall, it sounds like your Portal site is handling failover correctly, but your load balancer isn’t configured to redirect traffic to Portal 2 when Portal 1 fails.

Just to clarify, in a Portal high-availability setup, the primary–standby model applies only to the underlying postgres database. Portal's underlying tomcat web servers still operate in an active–active configuration, meaning both machines can receive and process web requests regardless of which node is primary or standby. Database read/writes occur on the primary postgres instance and are then replicated to the standby.

This is potentially why your configuration works when traffic is being routed to Portal 1 (standby), but Portal 2 is the designated primary.