我们可以打印 MLflow 服务器已启动的配置吗？

Question

I am using the following command to start the MLflow server:

mlflow server --backend-store-uri postgresql://mlflow_user:mlflow@localhost/mlflow  --artifacts-destination <S3 bucket location> --serve-artifacts  -h 0.0.0.0 -p 8000

Before production deployment, we have a requirement that we need to print or fetch the under what configurations the server is running. For example, the above command uses localhost postgres connection and S3 bucket.

Is there a way to achieve this?

Also, how do I set the server's environment as "production"? So finally I should see a log like this:

[LOG] Started MLflow server:
Env: production
postgres: localhost:5432
S3: <S3 bucket path>

Answer 1

You can wrap it in a bash script or in a Makefile script, eg

start_mlflow_production_server:
    @echo "Started MLflow server:"
    @echo "Env: production"
    @echo "postgres: localhost:5432"
    @echo "S3: <S3 bucket path>"
    @mlflow server --backend-store-uri postgresql://mlflow_user:mlflow@localhost/mlflow  --artifacts-destination <S3 bucket location> --serve-artifacts  -h 0.0.0.0 -p 8000

Additionally, it you can set and use environment variables specific to that server and print and use those in the command.