feat: update EvalHub and MLflow to match generated classes from openshift-python-wrapper#2691

sheltoncyril · sheltoncyril · commit 7ba826009f92 · 2026-04-06T20:46:40.000+01:00
Update resource classes to match the upstream generated code:
- EvalHub: add collections and otel fields, change providers type to list[Any]
- MLflow: change from NamespacedResource to Resource (cluster-scoped),
  add all new spec fields, remove namespace param from fixture

Signed-off-by: Shelton Cyril &lt;sheltoncyril@gmail.com&gt;
diff --git a/tests/model_explainability/evalhub/conftest.py b/tests/model_explainability/evalhub/conftest.py
@@ -106,7 +106,6 @@ def mlflow_instance(
     with MLflow(
         client=admin_client,
         name="mlflow",
-        namespace=py_config["applications_namespace"],
         storage={
             "accessModes": ["ReadWriteOnce"],
             "resources": {"requests": {"storage": "10Gi"}},
diff --git a/utilities/resources/evalhub.py b/utilities/resources/evalhub.py
@@ -15,27 +15,44 @@ class EvalHub(NamespacedResource):
 
     def __init__(
         self,
+        collections: list[Any] | None = None,
         database: dict[str, Any] | None = None,
         env: list[Any] | None = None,
-        providers: list[str] | None = None,
+        otel: dict[str, Any] | None = None,
+        providers: list[Any] | None = None,
         replicas: int | None = None,
         **kwargs: Any,
     ) -> None:
         r"""
         Args:
-            database (dict[str, Any]): Database configuration for the eval-hub service
+            collections (list[Any]): Collections is the list of OOTB collection names to mount into the
+              deployment. Each name must match a collection-name label on a
+              ConfigMap in the operator namespace.
+
+            database (dict[str, Any]): Database configuration for persistent storage. This field is required:
+              the operator will not start the service without an explicit
+              database configuration. Set type to "postgresql" with a secret
+              reference, or "sqlite" for lightweight/development deployments.
 
             env (list[Any]): Environment variables for the eval-hub container
 
-            providers (list[str]): List of evaluation providers to enable
+            otel (dict[str, Any]): OpenTelemetry configuration for observability. When set, the operator
+              includes OTEL settings in the generated config. When omitted, the
+              service uses its defaults (OTEL disabled).
+
+            providers (list[Any]): Providers is the list of OOTB provider names to mount into the
+              deployment. Each name must match a provider-name label on a
+              ConfigMap in the operator namespace.
 
             replicas (int): Number of replicas for the eval-hub deployment
 
         """
         super().__init__(**kwargs)
 
+        self.collections = collections
         self.database = database
         self.env = env
+        self.otel = otel
         self.providers = providers
         self.replicas = replicas
 
@@ -47,12 +64,18 @@ def to_dict(self) -> None:
             self.res["spec"] = {}
             _spec = self.res["spec"]
 
+            if self.collections is not None:
+                _spec["collections"] = self.collections
+
             if self.database is not None:
                 _spec["database"] = self.database
 
             if self.env is not None:
                 _spec["env"] = self.env
 
+            if self.otel is not None:
+                _spec["otel"] = self.otel
+
             if self.providers is not None:
                 _spec["providers"] = self.providers
 
diff --git a/utilities/resources/mlflow.py b/utilities/resources/mlflow.py
@@ -1,48 +1,288 @@
+# Generated using https://github.com/RedHatQE/openshift-python-wrapper/blob/main/scripts/resource/README.md
+
+
 from typing import Any
 
-from ocp_resources.resource import NamespacedResource
+from ocp_resources.resource import Resource
 
 
-class MLflow(NamespacedResource):
-    """MLflow is the Schema for the mlflows API."""
+class MLflow(Resource):
+    """
+    MLflow is the Schema for the mlflows API
+    """
 
     api_group: str = "mlflow.opendatahub.io"
 
     def __init__(
         self,
-        storage: dict[str, Any] | None = None,
-        backend_store_uri: str | None = None,
+        affinity: dict[str, Any] | None = None,
         artifacts_destination: str | None = None,
-        serve_artifacts: bool | None = None,
+        backend_store_uri: str | None = None,
+        backend_store_uri_from: dict[str, Any] | None = None,
+        ca_bundle_config_map: dict[str, Any] | None = None,
+        default_artifact_root: str | None = None,
+        env: list[Any] | None = None,
+        env_from: list[Any] | None = None,
+        extra_allowed_origins: list[Any] | None = None,
         image: dict[str, Any] | None = None,
+        network_policy_additional_egress_rules: list[Any] | None = None,
+        node_selector: dict[str, Any] | None = None,
+        pod_annotations: dict[str, Any] | None = None,
+        pod_labels: dict[str, Any] | None = None,
+        pod_security_context: dict[str, Any] | None = None,
+        registry_store_uri: str | None = None,
+        registry_store_uri_from: dict[str, Any] | None = None,
+        replicas: int | None = None,
+        resources: dict[str, Any] | None = None,
+        security_context: dict[str, Any] | None = None,
+        serve_artifacts: bool | None = None,
+        service_account_name: str | None = None,
+        storage: dict[str, Any] | None = None,
+        tolerations: list[Any] | None = None,
+        workers: int | None = None,
         **kwargs: Any,
     ) -> None:
+        r"""
+        Args:
+            affinity (dict[str, Any]): Affinity specifies the pod's scheduling constraints
+
+            artifacts_destination (str): ArtifactsDestination is the server-side destination for MLflow
+              artifacts (models, plots, files). This setting only applies when
+              ServeArtifacts is enabled. When ServeArtifacts is disabled, this
+              field is ignored and clients access artifact storage directly.
+              Supported schemes: file://, s3://, gs://, wasbs://, hdfs://, etc.
+              Examples:   - "file:///mlflow/artifacts" (requires Storage to be
+              configured)   - "s3://my-bucket/mlflow/artifacts" (no Storage
+              needed)   - "gs://my-bucket/mlflow/artifacts" (no Storage needed)
+              If not specified when ServeArtifacts is enabled, defaults to
+              "file:///mlflow/artifacts"  For cloud storage authentication, use
+              EnvFrom to inject credentials from secrets or configmaps. Example
+              for S3:   envFrom:   - secretRef:       name: aws-credentials  #
+              Contains AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY Example for GCS:
+              envFrom:   - secretRef:       name: gcp-credentials  # Contains
+              GOOGLE_APPLICATION_CREDENTIALS path
+
+            backend_store_uri (str): BackendStoreURI is the URI for the MLflow backend store (metadata).
+              Supported schemes: file://, sqlite://, mysql://, postgresql://,
+              etc. Examples:   - "sqlite:////mlflow/mlflow.db" (requires Storage
+              to be configured) Note: For URIs containing credentials, prefer
+              using BackendStoreURIFrom for security. This must be set
+              explicitly unless BackendStoreURIFrom is provided.
+
+            backend_store_uri_from (dict[str, Any]): BackendStoreURIFrom is a reference to a secret containing
+              the backend
+              store URI. Use this instead of BackendStoreURI when the URI
+              contains credentials. Mutually exclusive with BackendStoreURI -
+              the API rejects specs that set both.
+
+            ca_bundle_config_map (dict[str, Any]): CABundleConfigMap specifies a ConfigMap containing a CA certificate
+              bundle. The bundle will be mounted into the MLflow container and
+              configured for use with TLS connections (e.g. PostgreSQL SSL, S3
+              with custom certificates).
+
+            default_artifact_root (str): DefaultArtifactRoot is the default artifact root path for MLflow runs
+              on the server. This is required when serveArtifacts is false.
+              Supported schemes: file://, s3://, gs://, wasbs://, hdfs://, etc.
+              Examples:   - "s3://my-bucket/mlflow/artifacts"   - "gs://my-
+              bucket/mlflow/artifacts"   - "file:///mlflow/artifacts"
+
+            env (list[Any]): Env is a list of environment variables to set in the MLflow container
+
+            env_from (list[Any]): EnvFrom is a list of sources to populate environment variables in the
+              MLflow container
+
+            extra_allowed_origins (list[Any]): ExtraAllowedOrigins is a list of additional origins to allow for CORS
+              requests. The operator preconfigures safe defaults including
+              Kubernetes service names, the data science gateway domain, and
+              localhost. Use this field to add additional origins beyond the
+              defaults. Each entry should be a full origin
+              (scheme://host[:port]), e.g. "https://my-app.example.com".
+
+            image (dict[str, Any]): Image specifies the MLflow container image. If not specified, use the
+              default image via the MLFLOW_IMAGE environment variable in the
+              operator.
+
+            network_policy_additional_egress_rules (list[Any]): NetworkPolicyAdditionalEgressRules specifies
+              additional egress rules
+              to append to the default NetworkPolicy. The default policy permits
+              DNS (53), HTTPS (443), Kubernetes API (6443), PostgreSQL (5432),
+              MySQL (3306), and S3-compatible storage (MinIO 9000, SeaweedFS
+              8333). Use this field when connecting to services on non-standard
+              ports or when destination restrictions are needed.
+
+            node_selector (dict[str, Any]): NodeSelector is a selector which must be true for the pod to fit on a
+              node
+
+            pod_annotations (dict[str, Any]): PodAnnotations are annotations to add only to the MLflow pod, not to
+              other resources. Use this for pod-specific annotations like
+              Prometheus scraping or sidecar configuration.
+
+            pod_labels (dict[str, Any]): PodLabels are labels to add only to the MLflow pod, not to other
+              resources. Use this for pod-specific labels like version,
+              component-specific metadata, etc. For labels that should be
+              applied to all resources (Service, Deployment, etc.), use
+              commonLabels in values.yaml.
+
+            pod_security_context (dict[str, Any]): PodSecurityContext specifies the security context for the MLflow pod
+
+            registry_store_uri (str): RegistryStoreURI is the URI for the MLflow registry store (model
+              registry metadata). Supported schemes: file://, sqlite://,
+              mysql://, postgresql://, etc. Examples:   -
+              "sqlite:////mlflow/mlflow.db" (requires Storage to be configured)
+              If omitted, defaults to the same value as backendStoreUri. Note:
+              For URIs containing credentials, prefer using RegistryStoreURIFrom
+              for security.
+
+            registry_store_uri_from (dict[str, Any]): RegistryStoreURIFrom is a reference to a secret containing the
+              registry store URI. Use this instead of RegistryStoreURI when the
+              URI contains credentials. Mutually exclusive with RegistryStoreURI
+              - the API rejects specs that set both.
+
+            replicas (int): Replicas is the number of MLflow pods to run
+
+            resources (dict[str, Any]): Resources specifies the compute resources for the MLflow container
+
+            security_context (dict[str, Any]): SecurityContext specifies the security context for the MLflow
+              container
+
+            serve_artifacts (bool): ServeArtifacts determines whether MLflow should serve artifacts. When
+              enabled, adds the --serve-artifacts flag to the MLflow server and
+              uses ArtifactsDestination to configure where artifacts are stored.
+              This allows clients to log and retrieve artifacts through the
+              MLflow server's REST API instead of directly accessing the
+              artifact storage. When disabled, ArtifactsDestination is ignored
+              and clients must have direct access to artifact storage.
+
+            service_account_name (str): ServiceAccountName is the name of the ServiceAccount to use for the
+              MLflow pod. If not specified, a default ServiceAccount will be
+              "mlflow-sa"
+
+            storage (dict[str, Any]): Storage specifies the persistent storage configuration using standard
+              PVC spec. Only required if using SQLite backend/registry stores or
+              file-based artifacts. Not needed when using remote storage (S3,
+              PostgreSQL, etc.). When omitted, no PVC will be created - ensure
+              backendStoreUri, registryStoreUri, and artifactsDestination point
+              to remote storage. Example:   storage:     accessModes:
+              ["ReadWriteOnce"]     resources:       requests:         storage:
+              10Gi     storageClassName: fast-ssd
+
+            tolerations (list[Any]): Tolerations are the pod's tolerations
+
+            workers (int): Workers is the number of uvicorn worker processes for the MLflow
+              server. Note: This is different from pod replicas. Each pod will
+              run this many worker processes. Defaults to 1. For high-traffic
+              deployments, consider increasing pod replicas instead.
+
+        """
         super().__init__(**kwargs)
 
-        self.storage = storage
-        self.backend_store_uri = backend_store_uri
+        self.affinity = affinity
         self.artifacts_destination = artifacts_destination
-        self.serve_artifacts = serve_artifacts
+        self.backend_store_uri = backend_store_uri
+        self.backend_store_uri_from = backend_store_uri_from
+        self.ca_bundle_config_map = ca_bundle_config_map
+        self.default_artifact_root = default_artifact_root
+        self.env = env
+        self.env_from = env_from
+        self.extra_allowed_origins = extra_allowed_origins
         self.image = image
+        self.network_policy_additional_egress_rules = network_policy_additional_egress_rules
+        self.node_selector = node_selector
+        self.pod_annotations = pod_annotations
+        self.pod_labels = pod_labels
+        self.pod_security_context = pod_security_context
+        self.registry_store_uri = registry_store_uri
+        self.registry_store_uri_from = registry_store_uri_from
+        self.replicas = replicas
+        self.resources = resources
+        self.security_context = security_context
+        self.serve_artifacts = serve_artifacts
+        self.service_account_name = service_account_name
+        self.storage = storage
+        self.tolerations = tolerations
+        self.workers = workers
 
     def to_dict(self) -> None:
+
         super().to_dict()
 
         if not self.kind_dict and not self.yaml_file:
             self.res["spec"] = {}
             _spec = self.res["spec"]
 
-            if self.storage is not None:
-                _spec["storage"] = self.storage
+            if self.affinity is not None:
+                _spec["affinity"] = self.affinity
+
+            if self.artifacts_destination is not None:
+                _spec["artifactsDestination"] = self.artifacts_destination
 
             if self.backend_store_uri is not None:
                 _spec["backendStoreUri"] = self.backend_store_uri
 
-            if self.artifacts_destination is not None:
-                _spec["artifactsDestination"] = self.artifacts_destination
+            if self.backend_store_uri_from is not None:
+                _spec["backendStoreUriFrom"] = self.backend_store_uri_from
 
-            if self.serve_artifacts is not None:
-                _spec["serveArtifacts"] = self.serve_artifacts
+            if self.ca_bundle_config_map is not None:
+                _spec["caBundleConfigMap"] = self.ca_bundle_config_map
+
+            if self.default_artifact_root is not None:
+                _spec["defaultArtifactRoot"] = self.default_artifact_root
+
+            if self.env is not None:
+                _spec["env"] = self.env
+
+            if self.env_from is not None:
+                _spec["envFrom"] = self.env_from
+
+            if self.extra_allowed_origins is not None:
+                _spec["extraAllowedOrigins"] = self.extra_allowed_origins
 
             if self.image is not None:
                 _spec["image"] = self.image
+
+            if self.network_policy_additional_egress_rules is not None:
+                _spec["networkPolicyAdditionalEgressRules"] = self.network_policy_additional_egress_rules
+
+            if self.node_selector is not None:
+                _spec["nodeSelector"] = self.node_selector
+
+            if self.pod_annotations is not None:
+                _spec["podAnnotations"] = self.pod_annotations
+
+            if self.pod_labels is not None:
+                _spec["podLabels"] = self.pod_labels
+
+            if self.pod_security_context is not None:
+                _spec["podSecurityContext"] = self.pod_security_context
+
+            if self.registry_store_uri is not None:
+                _spec["registryStoreUri"] = self.registry_store_uri
+
+            if self.registry_store_uri_from is not None:
+                _spec["registryStoreUriFrom"] = self.registry_store_uri_from
+
+            if self.replicas is not None:
+                _spec["replicas"] = self.replicas
+
+            if self.resources is not None:
+                _spec["resources"] = self.resources
+
+            if self.security_context is not None:
+                _spec["securityContext"] = self.security_context
+
+            if self.serve_artifacts is not None:
+                _spec["serveArtifacts"] = self.serve_artifacts
+
+            if self.service_account_name is not None:
+                _spec["serviceAccountName"] = self.service_account_name
+
+            if self.storage is not None:
+                _spec["storage"] = self.storage
+
+            if self.tolerations is not None:
+                _spec["tolerations"] = self.tolerations
+
+            if self.workers is not None:
+                _spec["workers"] = self.workers
+
+    # End of generated code
