From 01dcb993c72e867b4121553b391dddea08ad9dd6 Mon Sep 17 00:00:00 2001 From: Simon Mo Date: Thu, 7 Jan 2021 08:57:24 -0800 Subject: [PATCH] [Serve] Rescale Serve's Long Running Test to Cluster Mode (#13247) Now that `HeadOnly` becomes the new default HTTP location, we can re-enable the long running tests to use local multi-clusters. (also fixed the controller's API to match up to date, we should have caught these, I will open issues for this.) --- release/long_running_tests/workloads/serve.py | 2 +- release/long_running_tests/workloads/serve_failure.py | 7 +++---- 2 files changed, 4 insertions(+), 5 deletions(-) diff --git a/release/long_running_tests/workloads/serve.py b/release/long_running_tests/workloads/serve.py index 59b230776..ae66e0e67 100644 --- a/release/long_running_tests/workloads/serve.py +++ b/release/long_running_tests/workloads/serve.py @@ -11,7 +11,7 @@ from ray.cluster_utils import Cluster num_redis_shards = 1 redis_max_memory = 10**8 object_store_memory = 10**8 -num_nodes = 1 +num_nodes = 4 cluster = Cluster() for i in range(num_nodes): cluster.add_node( diff --git a/release/long_running_tests/workloads/serve_failure.py b/release/long_running_tests/workloads/serve_failure.py index 534dcbda7..943826365 100644 --- a/release/long_running_tests/workloads/serve_failure.py +++ b/release/long_running_tests/workloads/serve_failure.py @@ -11,7 +11,7 @@ from ray.cluster_utils import Cluster num_redis_shards = 1 redis_max_memory = 10**8 object_store_memory = 10**8 -num_nodes = 1 +num_nodes = 4 cpus_per_node = 10 cluster = Cluster() for i in range(num_nodes): @@ -39,10 +39,9 @@ class RandomKiller: def _get_all_serve_actors(self): controller = self.client._controller - routers = list(ray.get(controller.get_routers.remote()).values()) + routers = list(ray.get(controller.get_http_proxies.remote()).values()) all_handles = routers + [controller] - worker_handle_dict = ray.get( - controller.get_all_worker_handles.remote()) + worker_handle_dict = ray.get(controller._all_replica_handles.remote()) for _, replica_dict in worker_handle_dict.items(): all_handles.extend(list(replica_dict.values()))