From a service management point of view, Service Fabric can be thought of as an orchestrator. An orchestrator in general terms is an automated piece of software used to manage service deployments. This piece of software is supposed to abstract the complexities around provisioning, deploying, fault handling, scaling, and optimizing the applications it is managing, from the end user. For instance, an orchestration should be able to consume a configuration which specifies the number of instances of service to run and perform the task of deploying the services-based on multiple complex factors such as resource availability on nodes in a cluster, placement constraints, and so on

Orchestrators are also responsible for fault handling and recovery of services. If a node in a cluster fails, the orchestrator needs to gracefully handle this while ensuring service availability. Updating a service deployment or applying a patch to the underlying framework is also managed by orchestrators. Apart from Service Fabric, there are other orchestration services available in the market, such as Mesosphere, Core OS, Docker Swarm, and Kubernetes. Azure container services open opportunities for some of these powerful orchestration services to be hosted on Azure. More about this is discussed later in this book.