About Parallel Method Invocation (PMI)
ScaleOut StateServer provides an in-memory compute engine that can invoke custom logic on the same server that hosts the ScaleOut service. When analyzing a large number of objects, this approach minimizes network overhead and takes advantage of data locality to reduce processing time.
Use Cases for Parallel Method Invocation
- Bulk updates: The modification or removal of large numbers of objects can be performed directly on the servers where the objects reside.
- Distributed aggregation: Large-scale distributed analysis can be performed, where a single aggregated result is returned to the client that invoked the PMI operation.