Fix: Project - allow retry (#2591) (#2594)

Co-authored-by: Ruth Netser <[email protected]>

Author: redhat-qe-bot

.github/workflows/code-check.yml.old

@@ -2,8 +2,11 @@
 
 from typing import Any
 
+from kubernetes.dynamic.exceptions import ForbiddenError
+
 from ocp_resources.project_project_openshift_io import Project
 from ocp_resources.resource import Resource
+from ocp_resources.utils.constants import DEFAULT_CLUSTER_RETRY_EXCEPTIONS, PROTOCOL_ERROR_EXCEPTION_DICT
 
 
 class ProjectRequest(Resource):
@@ -54,7 +57,14 @@ def deploy(self, wait: bool = False) -> Project:  # type: ignore[override]
             teardown=self.teardown,
             delete_timeout=self.delete_timeout,
         )
-        project.wait_for_status(status=project.Status.ACTIVE)
+
+        # When a ProjectRequest is created, the project is created and the user is added to the project.
+        # RBAC binding may not have propagated yet, causing transient 403 Forbidden errors, so we need to retry
+        # on ForbiddenError as well.
+        project.wait_for_status(
+            status=project.Status.ACTIVE,
+            exceptions_dict=PROTOCOL_ERROR_EXCEPTION_DICT | DEFAULT_CLUSTER_RETRY_EXCEPTIONS | {ForbiddenError: []},
+        )
 
         return project
 

ocp_resources/project_request.py

@@ -953,7 +953,13 @@ def _kube_v1_api(self) -> kubernetes.client.CoreV1Api:
         return kubernetes.client.CoreV1Api(api_client=self.client.client)
 
     def wait_for_status(
-        self, status: str, timeout: int = TIMEOUT_4MINUTES, stop_status: str | None = None, sleep: int = 1
+        self,
+        status: str,
+        timeout: int = TIMEOUT_4MINUTES,
+        stop_status: str | None = None,
+        sleep: int = 1,
+        exceptions_dict: dict[type[Exception], list[str]] = PROTOCOL_ERROR_EXCEPTION_DICT
+        | DEFAULT_CLUSTER_RETRY_EXCEPTIONS,
     ) -> None:
         """
         Wait for resource to be in status
@@ -962,6 +968,7 @@ def wait_for_status(
             status (str): Expected status.
             timeout (int): Time to wait for the resource.
             stop_status (str): Status which should stop the wait and failed.
+            exceptions_dict (dict[type[Exception], list[str]]): Dictionary of exceptions to retry on.
 
         Raises:
             TimeoutExpiredError: If resource in not in desire status.
@@ -971,10 +978,7 @@ def wait_for_status(
         samples = TimeoutSampler(
             wait_timeout=timeout,
             sleep=sleep,
-            exceptions_dict={
-                **PROTOCOL_ERROR_EXCEPTION_DICT,
-                **DEFAULT_CLUSTER_RETRY_EXCEPTIONS,
-            },
+            exceptions_dict=exceptions_dict,
             func=lambda: self.exists,
         )
         current_status = None