[hitnik]

Summary / Problem

During test cleanup, stopping or killing containers (such as tr-ct-dhcp-2 or tr-ui) when they have already exited or stopped caused Docker to raise HTTP 409 Conflict (Container ... is not running). This resulted in erroneous error logging ("Container stop error") during standard cleanup.
Linked Buganizer issue: b/496885581 ("[TR] Error during test cleanup: Attempting to kill non-running container tr-ct-dhcp-2").

Root Cause

1. docker_module.py and network_validator.py did not check container.status == 'running' before calling container.stop() or container.kill(), or did not handle docker.errors.APIError gracefully when status code 409/304 was returned.
2. In testrun.py, _stop_ui unconditionally called container.kill() without verifying whether the container was running, and caught only docker.errors.NotFound instead of also handling docker.errors.APIError.

Changes Made

1. framework/python/src/core/docker/docker_module.py:

   • Added import docker.errors.
   • In stop(self, kill=False):
     • Check if container.status == 'running': before stopping/killing.
     • If not running, log at DEBUG: Container <container_name> is not running (status: <status>).
     • Caught docker.errors.NotFound and log at DEBUG.
     • Caught docker.errors.APIError as error: if error.status_code in (304, 409) or 'not running' in str(error).lower(), log at DEBUG: Container <container_name> is already stopped: <error>, otherwise log at ERROR.

2. framework/python/src/net_orc/network_validator.py:

   • Added import docker.errors.
   • In _stop_network_device(self, net_device, kill=False):
     • Check if container.status == 'running': before stopping/killing.
     • If not running, log at DEBUG: Container <container_name> is not running (status: <status>).
     • Caught docker.errors.NotFound and log at DEBUG.
     • Caught docker.errors.APIError as e: if e.status_code in (304, 409) or 'not running' in str(e).lower(), log at DEBUG: Container <container_name> is already stopped: <e>, otherwise log at ERROR.

3. framework/python/src/core/testrun.py:

   • In _stop_ui(self):
     • Check if container.status == 'running': before calling container.kill().
     • In the outer try/except, catch (docker.errors.NotFound, docker.errors.APIError): pass.

Verification Plan

1. Run Testrun module execution and cleanup with containers that exit normally before cleanup (such as DHCP validator/testing containers).
2. Verify that stopping an already stopped/exited container logs a debug message rather than an ERROR.
3. Verify that _stop_ui exits cleanly whether tr-ui is running, exited, or not found.
4. Run automated test suite: pytest testing/unit/.

---

Branch comparison against dev: dev...fix-ai/container-stop-cleanup-496885581