Mariusz Strzelecki
Mariusz Strzelecki
This PR adds support of executing Kedro pipeline steps in AWS Fargate. No tests added yet :see_no_evil: --- Keep in mind: - [x] Documentation updates - [x] [Changelog](CHANGELOG.md) updates -...
Right now it's not possible to override parameters when submitting a new run. It makes it harder to do operations like airflow-oriented hyperparameter tuning (that would involve running many runs...
When the pipeline is executed, every node finishing sets the status in Mlflow to "SUCEEDED". Therefore, the status in Mlflow doesn't reflect status of last pipeline, only the status of...
Currently, the plugin doesn't allow to set affinity on ContainerOp. In our setup, with multitenant Kubeflow deployment, we need to separate physically pipelines started by different teams (by running them...
Pipeline generator [sanitizes the node name](https://github.com/getindata/kedro-kubeflow/blob/0.7.4/kedro_kubeflow/generators/pod_per_node_pipeline_generator.py#L127) to be kubeflow-compatible, but when it comes to details selection (like resources, tolerations), it doesn't follow the same logic, it purely checks for op...
When a volume is defined in the configuration, first step of the pipeline creates one and on-exit-handler removes it. However, it looks that since Nov 2021 there is way more...
When running `run-once` with invalid pipeline name, the error looks like: ``` File "/opt/conda/envs/python38/lib/python3.8/site-packages/kfp/compiler/compiler.py", line 1108, in _create_and_write_workflow workflow = self._create_workflow( File "/opt/conda/envs/python38/lib/python3.8/site-packages/kfp/compiler/compiler.py", line 896, in _create_workflow pipeline_func(*args_list, **kwargs_dict) File...
#### Description Provides alternative pipeline generation, suitable to run SparkML pipelines with lazy evaluation of steps and `MemoryDataSet` artifacts. ##### PR Checklist - [ ] Tests added - [ ]...