def check_run_worker_health(self, run: PipelineRun):
container = self._get_container(run)
if container == None:
return CheckRunHealthResult(WorkerStatus.NOT_FOUND)
if container.status == "running":
return CheckRunHealthResult(WorkerStatus.RUNNING)
return CheckRunHealthResult(
WorkerStatus.FAILED, msg=f"Container status is {container.status}")
def check_run_worker_health(self, run: PipelineRun):
job_name = get_job_name_from_run_id(
run.run_id, resume_attempt_number=self._instance.count_resume_run_attempts(run.run_id)
)
try:
job = self._batch_api.read_namespaced_job(namespace=self.job_namespace, name=job_name)
except Exception:
return CheckRunHealthResult(
WorkerStatus.UNKNOWN, str(serializable_error_info_from_exc_info(sys.exc_info()))
)
if job.status.failed:
return CheckRunHealthResult(WorkerStatus.FAILED, "K8s job failed")
return CheckRunHealthResult(WorkerStatus.RUNNING)
def check_run_worker_health(self, run: PipelineRun):
job_namespace = _get_validated_celery_k8s_executor_config(run.run_config).get(
"job_namespace"
)
job_name = get_job_name_from_run_id(run.run_id)
try:
job = self._batch_api.read_namespaced_job(namespace=job_namespace, name=job_name)
except Exception:
return CheckRunHealthResult(
WorkerStatus.UNKNOWN, str(serializable_error_info_from_exc_info(sys.exc_info()))
)
if job.status.failed:
return CheckRunHealthResult(WorkerStatus.FAILED, "K8s job failed")
return CheckRunHealthResult(WorkerStatus.RUNNING)