Summary

Changes the API of the event fetchers to fetch batch of events (max 100 for now) from a position

Summary

Implement faster event handlers by using the in-memory event log subscription in the filter stream processor so that it can process batches of events in memory

Summary

Adds functionality to subscribe to the event log

Summary

Adds functionality to create bootstrap procedures

Summary

Start integrating the Actor Model into the Runtime code with first introducing Proto.Actor into the Event Store. We're taking advantage of the Actor Model to implement batching logic on the write side of the Event Store (committing events) to massively increase the throughput for committing events. This is also built in a way that enables us to in the future be able to scale horizontally through having multiple Runtimes running together in clusters.

In addition we now do benchmarking on key parts of the system and publishing the results on each GitHub release so that we can keep an eye on the current performance.

Changed

• When committing events the commits will be written and responded in batches which in turn increases the throughput of commits by several times.

Summary

Adds more integration tests for the event store

Summary

Adds integration tests

Summary

Benchmarks for persisting and getting stream processor states

Summary

Increase number of events to process in projection benchmark

Summary

Projection benchmark