| commit | 2e9fdb6f59edeb6c78f9983e9da18ca743652ba7 | [log] [tgz] |
|---|---|---|
| author | Sanjeev Kulkarni <sanjeevrk@gmail.com> | Mon Jun 26 12:39:59 2017 -0700 |
| committer | Runhang Li <rli@twitter.com> | Tue Jul 18 13:36:52 2017 -0700 |
| tree | a0f74cd49b10898b437017bd316eb4e7082cd206 | |
| parent | 5dd673713923e8e1612c69018188dbd57636b064 [diff] |
Add support for src task id in the storm api (#1938) * Added src task id information of tuples flowing between stmgrs and between stmgrs and instance * StreamManager changes to support src_task_id. We now break tuple batches based on src_task_ids * Heron-instance changes to support src_task_id. It can pass the compliation. But unit tests will fail. Heron-simulator has not been changed yet. So src_task_id will not work in heron-simulator. * Fix a NPE in heron-instance * Added the support of src_task_id in heron-simulator. Fixed all unit tests and coding style. * Set src_task_id in python instance as well
Heron is realtime analytics platform developed by Twitter. It is the direct successor of Apache Storm, built to be backwards compatible with Storm's topology API but with a wide array of architectural improvements.
For more information:
We recently merged updates to run Heron natively using Mesos in AWS, Mesos/Aurora in AWS, and locally on a laptop.
We also added beta testing for Apache YARN support using Apache REEF, in addition to the Simple Linux Utility for Resource Management (SLURM).
We are working to add support for Mesosphere DC/OS and Kubernetes. We will continue to post updates as we progress.
The official documentation for Heron is located at heronstreaming.io. To contribute to documentation, build and run the documentation locally. More information can be found in the documentation README.