Fabric keeper. The fabric keeper exists on all workers to coordinate the distribution of new applications, application and configuration file updates, fabric system updates, and operating system updates. When fabric keeper receives instructions to perform an update, fabric keeper coordinates a rolling update. The fabric keeper tier is key to the scale-out virtualization and automated management features of the fabric.

Fabric worker manager. The fabric worker manager exists on all workers to monitor fabric services and self-organizing processing regions. If a fabric service stops running for any reason other than a manual stop, the fabric worker manager restarts that service. The self-organizing and self-healing functionality of the fabric worker manager tier supports the fabric's scale-out virtualization and application-level fault tolerance features.

Request processing services. The request processing services consist of Request services, Process services, and Task services. These services exist on all workers, and cooperate to process a request and return the result to the client. Working together, these services provide application-level fault tolerance and scale-out virtualization in the fabric.

Fabric accessible memory. The Fabric accessible memory (FAM) is fabric-based storage for applications to maintain and share data. The data physically exists in a cache of RAM located on multiple computers within the fabric rather than in a database tier. The data is available to any worker within the fabric.