All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
NOTE: For CLI interfaces, we support SemVer approach. However, for API components we don't use SemVer as of now. This may lead to instability when using dbx API methods directly.
Please read through the Keep a Changelog (~5min).
- Provided bugfix for emoji-based messages in certain shell environments
Unreleased changes must be tracked above this line. When releasing, Copy the changelog to below this line, with proper version and date. And empty the [Unreleased] section above.
- Add support for named property of the driver instance pool name
- Add support for built-in templates and project initialization via :code:
dbx init
- Provided bugfix for named property resolution in multitask-based jobs
- Update the contribution docs with CLA
- Update documentation about environment variables
- Add support for named job properties
- Add support for
spark_jar_task
in Azure Data Factory reflector
- Provide bugfix for strict path resolving in the execute command
- Provide bugfix for Azure Datafactory when using
existing_cluster_id
- Update
databricks-cli
dependency to 0.16.2 - Improved code coverage
- Added support for environment variables in deployment files
- Fixed minor bug in exception text
- Provide a bugfix for execute issue
- Removed pydash from package dependencies, as it is not used. Still need it as a dev-requirement.
- Added support for multitask jobs.
- Added more explanations around DATABRICKS_HOST exception during API client initialization
- Add strict path adjustment policy and FUSE-based path adjustment
- Fix issue which stripped non-pyspark libraries from a requirements file during deploys.
- Fix issue which didn't update local package during remote execution.
- Support for yaml-based deployment files.
- Now dbx finds the git branch name from any subdirectory in the repository.
- Minor alterations in the documentation.
- Altered the Changelog based on Keep a Changelog
- Changed(for contributors): Makefile now requires pyenv.
- Changed(for contributors): Makefile is more self describing and self-sufficient.
make clean install
will set you up with all that is needed.make help
to see all available commands.
- Fix issue with execute parameters passing
- Fix issue with multi-version package upload
- Add explicit exception for artifact location change
- Add experimental support for fixed properties' propagation from cluster policies
- Added Run Submit API support.
- Fixed the issue with pywin32 installation for Azure imports on win platforms.
- Integration with Azure Data Factory.
- Some small internal behaviour fixes.
- Changed the behaviour of
dbx deploy --write-specs-to-file
, to make the structure of specs file compatible with environment structure.
- Added integrated permission management, please refer to documentation for details.
- Added
--write-specs-to-file
option fordbx deploy
command.
- HotFix for execute command.
- Made Internal refactorings after code coverage analysis.
- Fixed issue with job spec adjustment.
- Finalized the CI setup for the project.
- No code changes were done.
- Release is required to start correct numeration in pypi.
- Initial public release version.