Skip to content

[CI] pass cloud into service.status calls #50860

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
aslonnie merged 2 commits into from
Feb 24, 2025
Merged

[CI] pass cloud into service.status calls #50860

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
dc2863c
[CI] pass cloud into service.status calls
GeneDer Feb 24, 2025
cfda08f
pass cloud to terminate command as well
GeneDer Feb 24, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions release/serve_tests/workloads/anyscale_service_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,8 +15,10 @@
logging.basicConfig(level=logging.INFO)


def check_service_state(service_name: str, expected_state: ServiceState):
state = service.status(name=service_name).state
def check_service_state(
service_name: str, expected_state: ServiceState, cloud: Optional[str] = None
):
state = service.status(name=service_name, cloud=cloud).state
logger.info(
f"Waiting for service {service_name} to be {expected_state}, currently {state}"
)
Expand All @@ -34,6 +36,7 @@ def start_service(
image_uri: Optional[str] = None,
working_dir: Optional[str] = None,
add_unique_suffix: bool = True,
cloud: Optional[str] = None,
):
"""Starts an Anyscale Service with the specified configs.

Expand Down Expand Up @@ -76,18 +79,20 @@ def start_service(
expected_state="RUNNING",
retry_interval_ms=10000, # 10s
timeout=600,
cloud=cloud,
)

yield service_name

finally:
logger.info(f"Terminating service {service_name}.")
service.terminate(name=service_name)
service.terminate(name=service_name, cloud=cloud)
wait_for_condition(
check_service_state,
service_name=service_name,
expected_state="TERMINATED",
retry_interval_ms=10000, # 10s
timeout=600,
cloud=cloud,
)
logger.info(f"Service '{service_name}' terminated successfully.")
6 changes: 4 additions & 2 deletions release/serve_tests/workloads/autoscaling_load_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,7 @@


URI = "https://serve-resnet-benchmark-data.s3.us-west-1.amazonaws.com/000000000019.jpeg"
CLOUD = "serve_release_tests_cloud"


@click.command()
Expand All @@ -47,7 +48,7 @@ def main(output_path: Optional[str], image_uri: Optional[str]):
],
}
compute_config = ComputeConfig(
cloud="serve_release_tests_cloud",
cloud=CLOUD,
head_node=HeadNodeConfig(instance_type="m5.8xlarge"),
worker_nodes=[
WorkerNodeGroupConfig(
Expand All @@ -67,9 +68,10 @@ def main(output_path: Optional[str], image_uri: Optional[str]):
compute_config=compute_config,
applications=[resnet_application],
working_dir="workloads",
cloud=CLOUD,
) as service_name:
ray.init(address="auto")
status = service.status(name=service_name)
status = service.status(name=service_name, cloud=CLOUD)

# Start the locust workload
num_locust_workers = int(ray.available_resources()["CPU"]) - 1
Expand Down
6 changes: 4 additions & 2 deletions release/serve_tests/workloads/replica_scalability.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@

DEFAULT_FULL_TEST_NUM_REPLICA = 1000
DEFAULT_FULL_TEST_TRIAL_LENGTH_S = 60
CLOUD = "serve_release_tests_cloud"


@click.command()
Expand Down Expand Up @@ -52,7 +53,7 @@ def main(
],
}
compute_config = ComputeConfig(
cloud="serve_release_tests_cloud",
cloud=CLOUD,
head_node=HeadNodeConfig(instance_type="m5.8xlarge"),
worker_nodes=[
WorkerNodeGroupConfig(
Expand Down Expand Up @@ -92,9 +93,10 @@ def main(
compute_config=compute_config,
applications=[noop_1k_application],
working_dir="workloads",
cloud=CLOUD,
) as service_name:
ray.init("auto")
status = service.status(name=service_name)
status = service.status(name=service_name, cloud=CLOUD)

# Start the locust workload
num_locust_workers = int(ray.available_resources()["CPU"]) - 1
Expand Down