- Allow dynamic_import to find the child-most class
- Allow kafka-rolling-restart script to be executed in clusters with SSL or SASL enabled (thanks dinegri!)
- Add --use-admin-client option to kafka-consumer-manager commands increased faster performance.
- Allow set_replication_factor command to work with topics that have mismatched replication factors in their partitions
- Add new preferred replica election command to generate json files for kafka-preferred-replica-election
- Allow for removing brokers in kafka-cluster-manager replace command
- Support consumer offset message value schema v3 (thanks mborst!)
- Add --active-controller-for-last flag to kafka-rolling-restart
- Use v1.4.7 of kafka-python to take advantage of performance upgrades
- Fix Travis build by unpinning Docker version
- Fix git tag
- Relax pytz pinned version in setup.py
- Fix configuration writing in kafka-manual-throttle
- Remove support for python3.4
- Fix iteration bug in _extract_actions_unique_topics with --max-movement-size
- Implement --head option for kafka-check command
- Fix bug with --max-movement-size default value causing stuck decommissions
- Relax requirements for PyYAML and pytz
- Use openjdk instead of oracle-java8 in acceptance tests
- Migrate to ubuntu 16.04 in acceptance tests
- Update team name and email
- Add new --force-progress flag in kafka-cluster-manager decommission
- Fix bug with moving 0-size partitions in decommission
- Fix bug with genetic balancer using size in decommission when --max-movement-size not specified
- Add new --max-movement-size and --auto-max-movement-size in kafka-cluster-manager decommission
- Upgrade dependency requirements to newer versions
- Fix bug with KafkaGroupReader acceptance test
- Deprecate zookeeper offset storage in kafka-consumer-manager
- Implement kafka-check replication_factor command
- Fix imports, pytest version and build failure due to flake8
- Add change to display generated reassignment plan even on failure of validation
- Fix bug in offset_get command that showed topics as unsubscribed if any partition had an offset of 0 committed
- Upgrade paramiko in kafka-utils (paramiko < 2.5.0)
- Fix bug in unsubscribe_topics command where all subscribed topics were being displayed as subject to change regardless of specified --topics
- Add new commands offset_set_timestamp and offsets_for_timestamp
- Add verbosity option for kafka-consumer-manager
- Bump setuptools version.
- Remove cryptography dependency
- Add tox extension
- Upgrade kafka-python version to 1.4.2
- Improve performance of broker decommission process
- kafka-check offline_partitions does not throw an exception for empty clusters anymore
- Add --broker-ids option to kafka-rolling-restart
- Fix offset_get and delete_group when no offsets folder under zk consumers node
- Add NoNodeError to get_brokers & get_topics
- Catch Exceptions for commands under empty cluster exception
- Set exactly topic matching as default for get_topic_watermark command
- Add -r/--regex option for get_topic_watermark command for regex search
- Improve handling of missing topic in getting topic-specific configuration
- Add --topics option to kafka_consumer_manager unsubscribe_topics
- List unavailable-brokers in case of unavailable-replicas
- Update CHANGELOG.rst for version 1.4.0
- Add fetching creation time of topic and partition from zookeeper
- Fix build for kafka 0.10
- Refresh ssh connection after post_stop task (simplesteph)
- Add ssh config support to kafka-rolling-restarat (stephane)
- Add custom start and stop command to kafka-rolling-restart (stephane)
- Fix documentation for offset_get command
- Add unhandled exception logging to kafka-cluster-manager
- Fix kafka-cluster-manager-argument
- Add partition count and leader count to genetic rebalancer criterias
- Add python3 support (kennydo)
- Remove fabric dependency and use paramiko (jparkie)
- Fix kafka topic config setter
- Add revoke-leadership feature in kafka-cluster-manager
- Bump kafka-python to 1.3.3
- Fix genetic balancer generation limit
- Bump version to fix v1.0.0 tagging issue
- Bump version to change command from under_replicated to replica_unavailability
- Refactor kafka group reader
- Support missing local cluster in config
- add generic prechecks in kafka-rolling-restart tool
- pin upper limit of kafka-python
- Fixes terminate for expection cases in kafka-check
- Optionally sort kafka-consumer-manager output by offset distance
- Support json output for kafka-checks
- kafka-python>=1.3.2,<1.4.0 in setup.py
- 0.10 integration tests
- Fix list_topics flakiness in kafka-consumer-manager
- Upgrade kafka-python in use to 1.3.2
- Use new KafkaConsumer for KafkaGroupReader
- Fix KafkaGroupreader when reading consumer group with partition zero.
- Add storage option for a few kafka_consumer_manager subcommands
- Change default offset storage from zookeeper to kafka
- Autodetecting the number of partitions for the __commit_offsets topic
- Fix integration tests
- Refactor kafka-cluster-manager to support multiple balancer classes and metrics
- Add PartitionMeasurer class and --partition-measurer option for providing user partition metrics
- Add --genetic-balancer option to kafka-cluster-manager to make use of the genetic balancer
- Change kafka-cluster-manager stats command output to include user partition metrics
- Add --show-stats option to kafka-cluster-manager rebalance
- Fetch group topics only from a single __consumer_offsets partition
- Add offline partitions check for kafka-check
- Fix set_replication_factor command plan generation
- Fix offset_get when the group name is stored only in kafka
- Add offset_set retry when writing offsets to kafka
- Fix a rebalance bug that would not generate a convergent assignment
- Check for pending asssignment before fetching the cluster topology
- Docs fixes
- Add short options from cluster-type and cluster-name
- Add option to see offset-distance for a consumer-group
- Add command set_replication_factor command
- Fix kafka-cluster-manager error on empty clusters
- Fix bug in cluster rebalance while updating sibling_distance
- Fix bug in cluster rebalance when replication group is None
- Add get topic watermark command
- Fix offset get json output
- Fix bug in decommissioning of failed brokers
- Make min_isr and under replicated partitions check much faster
- Use error field from metadata response in under replicated partition check
- Fix small typo in cluster manager logging
- Refactor under replicated partition check to use metadata request
- Add minimum replica number parameter to under replicated check
- Fix cluster manager logging
- Add verbose option to kafka-check
- Add under replicated partition check
- Add log segment corruption check
- Fix decommission command bug that caused decommission to fail in some cases
- Fix config when HOME env variable is not defined
- Fix bug for no available under-loaded brokers
- Fix group-parser local import
- Initial open-source release