With NebulaStream you can use all your computing resources, within or outside the cloud, to participate in the distributed processing of your queries. The only requirements are that the processing device is capable of running a NebulaStream worker and that the node has a connection to the coordinator. If both requirements are fulfilled, all your hardware can contribute.

Available Features

In the current release, we support a basic set of features:

  • Submitting multiple queries and merging them into one query plan.
  • Applying a basic set of rewriting and optimization rules.
  • Deploying queries in a distributed topology.
  • Processing multiple queries on the same node and across different nodes.
  • Compiling queries to efficient C++ code.
  • Basic complex event processing functionality.
  • Basic monitoring capabilities.
  • Execution on ARM and Intel CPUs.


We plan to add the following features in the near future:

  • Basic fault tolerance mechanisms.
  • Spatial processing capabilities.
  • User-defined functions specified in different programming languages (e.g., Python, Java, etc.).
  • ML-driven operator placement.
  • Adaptive sensor management.