Release Notes for Apache Helix 1.3.1
The Apache Helix team would like to announce the release of Apache Helix 1.3.1.
This is the tweenty nighth release under the Apache umbrella, and the tweenty fifth as a top-level project.
Helix is a generic cluster management framework used for the automatic management of partitioned, replicated and distributed resources hosted on a cluster of nodes.
In this release, Helix provides several performance improvements for rebalance pipeline:
Key Note for Helix Release
Helix Cloud Support
- Support better auto registration logic
- Can smoothly evacuate the node partitions with n -> n + 1 -> n strategy to other nodes.
Detailed Changes
Bug Fixes
- Fix the 'BadVersion' exception handling in doSet() code path. (#2614)
- Changed Log level to DEBUG for the top-state metrics logs, to avoid logging for every partition.(#2611)
- Fix inconsistent map key formatting in helix-rest maintenance API (#2583)
- WAGED - bugs fixes in n - n + 1 (#2578)
Improvements
- Implement the on-demand rebalance service (#2595)
- Metaclient leader election -(#2558, #2560, #2562)
- MetaClient Distributed Semaphore API and Permit object (#2520)
- ZkClient - only register one time watcher for read data when not using persist listener. (#2555)
- Add InstanceConfigs inputField to allow overriding the existing InstanceConfigs for /partitionAssignment API.(#2586)