The Apache Helix team would like to announce the release of Apache Helix 0.8.2.

This is the fourteenth release under the Apache umbrella, and the tenth 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:

Detailed Changes

New Features

  • [ROUTER] Expose ExternalViews in RoutingTable and RoutingTableSnapshot
  • [TASK] Expose UserContentStore in TaskDriver
  • [TASK] Add ThreadCountBasedAssignmentCalculator and integrate with Workflow/JobRebalancer and fix rebalancing logic
  • Should not start periodical refresh in RoutingTableProvider if helixmanager is not set.
  • Allow downward state transition during recovery and add recovery threshold
  • Support batch add jobs to JobQueue
  • Add periodic refresh to RoutingTableProvider

Bug Fixes

  • Make swap instance more robust and idempotent
  • Fix the issue that resource MBean may not be cleaned up when the resource is dropped.
  • Check NPE in getInstancesInClusterWithTag and throw more meaningful exception
  • Fix GroupCommit issue for adding back current state
  • Fix zkclient state related check and some potential bugs.
  • Fix race condition in Helix messaging service.
  • Fix bugs in ClusterStateVerifier and StrictMatchExternalViewVerifier.
  • Verify downward load balance and fix TestPartitionMovementThrottle


  • Adding RoutingTableProvider monitor for tracking refresh actions and routing table update callbacks.
  • Helix manager should support getting metadata store connection string
  • Add missing replica partition metric into ClusterStatus.
  • Improve BestPossibleExternalViewVerifier by avoid fully refreshing entire cluster data everytime and avoid to calulate bestpossible states for not required resources.
  • Support event differentiated logging
  • Helix job should fail immediately if the target resource is disabled
  • Make ZkClient keep retrying connect on expiring.
  • DelayedAutoRebalancer should calculate assignment based on full partition list.
  • Change RoutingDataCache to use zk version based selective update when reading the ExternalViews and TargetExternalView.
  • Change migration strategy to N -> N+1 -> N model
  • Add ThreadPoolExecutor purge and shutdown to avoid memory leak in RoutingTableProvider

Cheers, -- The Apache Helix Team

Back to top

Reflow Maven skin by Andrius Velykis.

Apache Helix, Apache, the Apache feather logo, and the Apache Helix project logos are trademarks of The Apache Software Foundation. All other marks mentioned may be trademarks or registered trademarks of their respective owners.
Privacy Policy