I see more and more customers, larger Enterprise ones in particular, running with multiple instances of MariaDB MaxScale. There are, as far as I can see, two somewhat different use cases.

• MaxScale HA - This is the basic case, used many customers, which is protect against MaxScale failure and to allow for uninterupted MaxScale upgrades. This works fine as it stands and is supported by cooporative monitoring and connector or firewall based failover.
• Full stack HA - This is the case when a customer has the full application stack, app, MariaDB Server and MaxScale, in a separate data center. Failover here is often a bit more intriguing and might involved DNS failover or manual failover. Often each DC has 2 MaxScale instances also as per above. There is no cooperative monitoring cross DC then and the main usecase for this setup is Disaster Recovery and similar. This can be seen in use at larger customers.

What I am proposing here is a solution to group and monitor a set of MaxScale instances from one MaxScale. In the second use case above, as it stands you would have 2 separet MaxScale and MariaDB Server setups that has little in common (except replication to the DR cluster). But even in the case of MaxScale HA as in the first case above, you would really like to see what is going on and how this MaxScale instance sees things, this might well be acive for example so any traffic in terms of connections and queries is the summary what is seen by the two MaxScale instances.

This could start out simple, for example a means of using the API on one of the MaxScale instances to get data from the other and some reasonable summary statistics, and then some of these summaries also presented in the GUI.

Exactly what measurements are presented and provided can be discussed, but this has the potential to be a neat Enterprise feature that would be easy to sell and talk about.