README.md

ML Flow implementation for GCP

Files overview

bucket_storage.tf - Creates bucket that will store Ml Flow artifacts. Secret is created to hold url to bucket.

cloud_run.tf - Holds configuration for the CloudRun implementation. The resources assigned are sufficient for the low traffic, cpu will be assigned only when the application is in use.

db_main.tf - DB instance using postgres (just instance itself database is created in separate file).

db_mlflow.tf - Data-base and its user, the user data (username and password) and url for database are placed in secrets as it will be used later.

iam.tf - IAM policy for database, secrets and bucket.

main.tf - Basic data about the project setup.

service_accounts.tf - Service account for the CloudRun and for objects creation in bucket.

variables.tf - Variables that need to be adjusted for individual project.

Requirements

Name	Version
terraform	>= 1.5
google	4.58.0

Resources

Name	Type
google_artifact_registry_repository.ar_mlflow	resource
google_cloud_run_service_iam_binding.access_to_mlflow	resource
google_cloud_run_v2_service.mlflow_on_cloudrun	resource
google_project_iam_member.cloudsql_connect	resource
google_secret_manager_secret.bucket_object_creator_account_key	resource
google_secret_manager_secret.db_login_ml_flow	resource
google_secret_manager_secret.db_password_ml_flow	resource
google_secret_manager_secret.mlflow_artifact_url	resource
google_secret_manager_secret.mlflow_database_url	resource
google_secret_manager_secret_iam_member.db_login_ml_flow_accessor	resource
google_secret_manager_secret_iam_member.db_password_ml_flow_accessor	resource
google_secret_manager_secret_iam_member.mlflow_artifact_url_accessor	resource
google_secret_manager_secret_iam_member.mlflow_database_url_accessor	resource
google_secret_manager_secret_version.ml-flow-artifact-url-version-basic	resource
google_secret_manager_secret_version.ml-flow-pass-version-basic	resource
google_secret_manager_secret_version.ml-flow-username-version-basic	resource
google_secret_manager_secret_version.mlflow_database_url-version-basic	resource
google_secret_manager_secret_version.mlflow_key-version-basic	resource
google_service_account.bucket_object_creator	resource
google_service_account.cloudrun-mlflow	resource
google_service_account_key.key_to_bucket_object_creator	resource
google_sql_database.ml_flow_db	resource
google_sql_database_instance.main	resource
google_sql_user.ml_flow_user	resource
google_storage_bucket.mlflow_artifacts_bucket	resource
google_storage_bucket_iam_policy.policy_bucket_object_create	resource
random_password.db_password_ml_flow	resource
google_iam_policy.mlflow-bucket-policy	data source
google_project.project	data source

Inputs

Name	Description	Type	Default	Required
bucket_users_list	List of users	list	[]	no
env	Name of the environment	string	"prod"	no
image_name	Name of the imagine that will be used.	string	"mlflow-imagine"	no
project_name	Full name of the project	string	"test-ml-flow-terraform"	no
region	GCP region that will be used for the project	string	"europe-west2"	no
vpn_to_access_db	VPN that will be used to connect to DB, while using 0.0.0.0/0 the application will be available from any IP (it will be accessible from the internet).	string	"0.0.0.0/0"	no
The inputs need to be updated with correct values before the project can be started.
Either in variables.tf or in terraform.tfvars.

Outputs

No outputs.