1.0.5

• kill tasks from a job (@ananaysingh )
• fetch all aurora & mesos master (@lawwong1 )
• list all tasks given a status (@lawwong1 )

1.0.4

• fetch free capacity
• simulate task fitting - compute how many tasks can be fit in the remaining capacity

1.0.3

• update CI to compile on PRs
• Add tier and production in task config
• Add fetch quota command
• Add priority into job config

1.0.2

• Bug fix: Bumped up backported thrift dependency to v0.13.2 since v0.13.1 was causing issues.
• Fetch mesos leader ability added (thanks @lenhattan86)
• Fall back on using Mesos agent to fetch Mesos leader if no zookeeper specified (thanks @zorro786).

1.0.1

Adding flag --monitor which is set to true by default to control whether the client verifies that the create or kill command are successful.

1.0.0

Now that Australis is able to launch rudimentary jobs, including launching jobs with Thermos support, as well as creating Job updates, it is not lacking any feature versus the python 2 client that is critical.

It should be noted that Australis and the python 2 client have different goals. Australis is not meant to be used for complex, templated style jobs. For such goals it's preferable to use gorealis and store the job configuration as code.

Australis is more of a Swiss army knife for Aurora Scheduler cluster administrators and scheduler developers who need to run quick jobs to test features.

0.22.0

• Support for starting job updates.
• Support for setting SlaAwareness for updates.
• Support for scheduling cron jobs.

0.1.1

• Updated dependencies.
• A patched version of the thrift 0.12.0 library to not panic when trying to write to a closed connection.
• First version released under the new home of australis on the aurora-scheduler repository.

0.1.0

• Adding support for drain sub-command to take in JSON list from stdin or from a specified file.
• Added flags json-file and json to drain, maintenance, and sla-drain.

0.0.9

• Added ability to create jobs which contain an executorless docker container.
• Sla-aware draining sub-command has been simplified. Instead of having a count/percentage
  subcommand, it now has a flag for each of these options. The count and percentage flag are
  mutually exclusive, and one of them has to be set.