- pull capacity report via /offers endpoint.
- calculate how many tasks (with resource and constraints) can be fit in the cluster.
examples of using the above 2 features are in aurora-scheduler/australis#33
* Remove unnecessary files from the thrift repository that come along with the go library.
* Updating thrift generated code to be 0.12.0 final generated code.
* Remove git.apache.org dependency in vendor folder.
* Migrating from git.apache.org/thrift.git to github.com/apache/thrift
* Upgrading dep (although it will not work now that imports are using mod format, it allows for users to easily fix this with a replacement of the import path).
* Upgrading mod dependencies for Thrift to point to github.com location of the repository.
* Bug fix for Thermos Payload generation relating to the GPU being set.
* Detecting if the transport error was not temporary in which case we stop retrying. Changed bug where get results was being called before we checked for an error.
* Adding exception for EOF error. All EOF errors will be retried.
* Addressing race conditions that may happen when client is closed or connection is re-established.
* Adding documentation about how this particular implemantion of the realis client uses retries in scenarios where a temporary error is found.
* Bugfix: switch statements were missing fallthrough statement thus making them retry non-retriable errors. Using a list to catch cases now.
* Adding tests for CreateService, createService when the executor doesn't exist, and createJob when the executor doesn't exist. Renamed Pulse test to reflect that it's using CreateService instead of CreateJob.
* Repsonse propagate back up to caller for context for CreateJob, CreateService, and StartJobUpdate.
* Deleting PR template as Travis CI takes care of running tests and formatting tests now.
* Change TaskQuery struct parameters to optional
* Thrift API is modified to make all the parameters in the
TaskQuery struct optional
* Autogenerated code is regenerated
* Changes in TaskQuery structs used in the project
* Now that TaskQuery receive optional values, pointers
instead of values must be passed to the struct
* Fixing possible race condition when passing backoff around as a pointer.
* Adding a debug logger that is turned off by default.
Info logger is enabled by default but prints out less information.
* Removing OK Aurora acknowledgment.
* Making Mutex a pointer so that there's no chance it can accidentally be copied.
* Changing %v to %+v for composite structs. Removing a repetitive statement for the Aurora return code.
* Removing another superflous debug statement.
* Removing a leftover helper function from before we changed how we configured the client.
* Changing the logging paradigm to only require a single logger. All logging will be disabled by default. If debug is enabled, and a logger has not been set, the library will default to printing all logging (INFO and DEBUG) to the stdout.
* Minor changes to demonstrate how a logger can be used in conjunction to debug mode.
* Removing port override as it is not needed
* Changing code comments to reflect getting rid of port override.
* Adding port override back in.
* Bug fix: Logger was being set to NOOP despite no logger being provided when debug mode is turned on.
* Turn on logging by default.
* Removing option to override schema and ports for information found on Zookeeper.
* Turning off debug mode for tests because it's too verbose. Making sure LevelLogger is initialized correctly under all scenarios.
* Removing override fields for zk config.
* Remove space.
* Removing info that is now incorrect about zk options.
* Fixing possible race condition when passing backoff around as a pointer.
* Adding a debug logger that is turned off by default. If debug is turned on, but a logger has not been assigned, a default logger that will print to STDOUT will be created.
* Making Mutex a pointer so that there's no chance it can accidentally be copied.
* Removing a leftover helper function from before we changed how we configured the client.
* Minor changes to demonstrate how a logger can be used in conjunction to debug mode in the sample client.
* ZK retries have been cleaned up. We will now retry after every error
EXCEPT when we have a badly formed path.
* ZK library has been reworked with optional arguments pattern to not be
so intertwined with the cluster.json file.
* Timeout error has been re-implemented as RetryError. RetryError
behaves like a Timeout error but is used exclusively to add more context
privately. This allows us to have unit tests that check our retry
mechanism is actually retrying.
* Additional logging has been added to retry mechanisms as well as to
the Zookeeper library we use.
