I just setup a Multi Region Wordpress Cluster on Jelastic from the marketplace. I have two environnement on two different region. My domain is pointing to a load balancer who is configured to send everything to server 1 and switch to the server 2 when server 1 crash/stop.
Everything works fine except when I shutdown server 1. The site is working but I can't access the admin (Error 500: Internal Server Error) and I figured that the problem appear also when I go to server 2 without the load balancer.
So I'm wondering if this is suppose to work like that ? Like the server 2 is only here to let the visitor have no downtime but you can't write anything in the database.
(I already tried to reinstall a similar environnement and I get the same issue)
Yes, such behaviour is expectable. This is caused by configuration of HA distributed shared storage based on GlusterFS. The only server1(master cluster) is an entry point for the writes and distributes content to the server2. As you probably has noted the Mutli-region cluster is on Beta and we're working on its improvment. In near future we will issue an add-on which will do reassigning the master role from one cluster to any other - in your case to delegate master role from server1 to server2 with a single click.