I am building an application with two processes that talk to the same database. They don't interact directly. I am setting them up to run in ECS fargate. Should I be using a single cluster with two services each pointing to a different task or should I have two different clusters? What do I need to consider when making that decision? Scaling requirements for both processes are going to be similar. They don't necessarily need to be isolated from each other.
Can I use two different task definitions within a single service