Configure a Local IDE using an SSH Gateway
The specifics for how to configure a local IDE to work with Cloudera Machine Learning are dependent on the local IDE you want to use.
Cloudera Machine Learning relies on the SSH functionality of the editors
to connect to the SSH endpoint on your local machine created with the
cdswctl client. Users establish an SSH endpoint on
their machine with the
cdswctl client. This
endpoint acts as the bridge that connects the editor on your machine and the Cloudera Machine
The following steps are a high-level description of the steps a user must complete:
- Establish an SSH endpoint with the CML CLI client. See https://docs.cloudera.com/machine-learning/1.0/cli/topics/ml-create-ssh-endpoint.html.
- Configure the local IDE to use Cloudera Machine Learning as the remote interpreter.
- Optionally, sync files with tools (like mutagen, SSHFS, or the functionality built into your IDE) from Cloudera Machine Learning to your local machine. Ensure that you adhere to IT policies.
- Edit the code in the local IDE and run the code interactively on Cloudera Machine Learning.
- Sync the files you edited locally to Cloudera Machine Learning.
- Use the Cloudera Machine Learning web UI to perform actions such as deploying a model that uses the code you edited.
You can see an end-to-end example for PyCharm configuration in the CML Editors Pycharm.