Skip to content

vscode-launch-json.md

Latest commit

 

History

History
128 lines (116 loc) · 4.25 KB

vscode-launch-json.md

File metadata and controls

128 lines (116 loc) · 4.25 KB

Visual Studio Code - launch.json sample

If you want to debug either the cluster_manager.py or the train_pipeline.py using Visual Studio Code, you can take advantage on this launch.json sample to run it directly from the IDE.

To use it, you need to fill in all the environment variables according to your cloud infrastructure setup.

Then, you can either add a launch.json with this content to your .vscode folder or you can merge an existing file content with the filled sample.

You will end up viewing something similar to this:

Local VSCode Debug Options

Tip

This is one sample of a launch.json that you can use on your Visual Studio Code. There are a variety of different approaches you can take to create your own launch.json to run this code locally.

Alternatively, consider running the shell script and then you can remove all of the env sections from the launch.json configuration file.

Feel free to submit a Pull Request with different samples of this file, so other options are provided to the community to have different experiences.

To understand each environment variable and how to get them, check the Project environment variables page.

To know more about debugging in Visual Studio Code, visit this link.

launch.json file structure

{
    "version": "0.2.0",
    "configurations": [
        {
            "name": "01 - Create a new Cluster",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "env": {
                "DATABRICKS_DOMAIN": "",
                "DATABRICKS_ACCESS_TOKEN": "",
                "DATABRICKS_CLUSTER_NAME_SUFFIX": "mySuffix"
            }
        },
        {
            "name": "02 - Terminate a Cluster",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "env": {
                "DATABRICKS_DOMAIN": "",
                "DATABRICKS_ACCESS_TOKEN": "",
                "DATABRICKS_CLUSTER_ID": ""
            },
            "args": [
                "--terminate"
            ]
        },
        {
            "name": "03 - Start an existing Cluster",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "env": {
                "DATABRICKS_DOMAIN": "",
                "DATABRICKS_ACCESS_TOKEN": "",
                "DATABRICKS_CLUSTER_ID": ""
            }
        },
        {
            "name": "04 - Terminate a Cluster (Permanently)",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "env": {
                "DATABRICKS_DOMAIN": "",
                "DATABRICKS_ACCESS_TOKEN": "",
                "DATABRICKS_CLUSTER_ID": ""
            },
            "args": [
                "--terminate",
                "--permanent"
            ]
        },
        {
            "name": "05 - Train model",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "env": {
                "DATABRICKS_DOMAIN": "",
                "DATABRICKS_ACCESS_TOKEN": "",
                "DATABRICKS_CLUSTER_ID": "",
                "AML_WORKSPACE_NAME": "MLOpsOSS-AML-WS",
                "RESOURCE_GROUP": "MLOpsOSS-AML-RG",
                "SUBSCRIPTION_ID": "",
                "TENANT_ID": "",
                "SP_APP_ID": "",
                "SP_APP_SECRET": "",
                "SOURCES_DIR": "",
                "TRAIN_SCRIPT_PATH": "src/train/train.py",
                "DATABRICKS_WORKSPACE_NAME": "MLOpsOSS-AML-ADB",
                "DATABRICKS_COMPUTE_NAME_AML": "ADB-Compute",
                "MODEL_DIR": "/dbfs/model",
                "MODEL_NAME": "mymodel"
            }
        }
    ]
}

Read more