- Fix a bug where an out of date configuration ID would be selected when restarting or replacing pods. This could lead to configuration updates being reverted to the values with which the service was initially deployed. (#2694)
hdfs.auth_to_local setting no longer need to be specified for the HDFS-specific principals when installing kerberized HDFS.
- Upgrade JRE to 1.8u192 to address CVEs
- Fix duplicate mounts being generated for TLS secrets, causing pod maintenance operations to fail (#2577)
- The SDK tests now validate missing values for
svc.yml Mustache variables. (#2527)
- Support for deploying the service in a remote region.
- Expose heap settings for all nodes.
- Support for HDFS rack awareness using DC/OS zones on DC/OS 1.11+
- Support for the automated provisioning of TLS artifacts to secure HDFS communication.
- Support for Kerberos authorization and authentication.
- Ability to pause a service pod for debugging and recovery purposes.
- Major Improvements to the stability and performance of service orchestration
- Upgrade JRE to 1.8u162. (#2135)
- The service now uses the Mesos V1 API. The service can be set back to the V0 API using the service property
- Placement constraints are now exposed.
- Dashes in envvars replaced with underscores to support Ubuntu.
- Some numeric configuration parameters could be interpreted incorrectly as floats, and are fixed.
- Uninstall now handles failed tasks correctly.
- Numerous fixes and enhancements to service reliability.
- Installation in folders is supported
- Use of a CNI network is supported
- Upgraded to dcos-commons 0.20.1
- Upgraded to
- Default user is now
- Allow configuration of scheduler log level
- Added a readiness check to journal nodes
- Pre-install notes include five agent pre-requisite
- Updated CLI documentation