Hi Jay,
When you say
do you mean I should be treating the two containers as a single site behind a sensible hostname and failover setup (i.e. same DISCOURSE_HOSTNAME with a load balancer/health checks in front), rather than exposing both containers’ ports directly?
I’m assuming both configs can just join the same Docker network rather than binding host ports, which would avoid some of the obvious conflicts — is that the right approach?
And for the volumes: section, do the bind mounts need to be different for each container (e.g. separate /var/discourse/shared/app1 and /var/discourse/shared/app2), or is there a sane way to share some of that between them?
Thanks!