I have raised this concern to my company associated appdynamic account managers last year but thought to share my views to see if anyone working/worked/informed on this requirement.
So my ask is to have continuous Availability of appdynamics controller with minimal outage or downtime for controller maintenance is required i.e. somewhat an active-active architecture.
Believe me - current appd architecture is too old school with watchdog and all. We have multiple deployments and each time we face some or the other issue. For e.g. we use to have hygiene infra updates and get the servers restarted each quarter breaking replication. Though we plan it out in such a way that either of nodes (primary/secondary) is active. But post which as soon as final replication is triggered, it demands for outage.
I know appd is working to re-architect the architecture but as per my info controller is out of scope i.e. controllers (esp app server of appdynamics) will still be HA and not CA.
I don't want to compare with other APM tools but I really want to say that lot of work needs to be done on appdynamics architecture to meet new design requirements (resilient, horizontal scalable, non-disruptive, continuous available, stable)
Hope it makes any sense.