diff --git a/release/serve_tests/workloads/serve_test_cluster_utils.py b/release/serve_tests/workloads/serve_test_cluster_utils.py
index 40cb852bb2a9..22e4e30cfdf3 100644
--- a/release/serve_tests/workloads/serve_test_cluster_utils.py
+++ b/release/serve_tests/workloads/serve_test_cluster_utils.py
@@ -9,7 +9,7 @@
 
 # Cluster setup configs
 NUM_CPU_PER_NODE = 10
-NUM_CONNECTIONS = 100
+NUM_CONNECTIONS = 10
 
 
 def setup_local_single_node_cluster(num_nodes):
diff --git a/release/serve_tests/workloads/serve_test_utils.py b/release/serve_tests/workloads/serve_test_utils.py
index 8c1db52c5e69..5f4d7d60e93b 100644
--- a/release/serve_tests/workloads/serve_test_utils.py
+++ b/release/serve_tests/workloads/serve_test_utils.py
@@ -200,7 +200,6 @@ def run_one_wrk_trial(trial_length: str,
             "-d",
             trial_length,
             "--latency",
-            "--timeout=5s",
             f"http://{http_host}:{http_port}/{endpoint}",
         ],
         stdout=PIPE,
diff --git a/release/serve_tests/workloads/single_deployment_1k_noop_replica.py b/release/serve_tests/workloads/single_deployment_1k_noop_replica.py
index df05fcc226b7..4aedd3fb7d63 100644
--- a/release/serve_tests/workloads/single_deployment_1k_noop_replica.py
+++ b/release/serve_tests/workloads/single_deployment_1k_noop_replica.py
@@ -39,7 +39,7 @@
 from serve_test_cluster_utils import (
     setup_local_single_node_cluster,
     setup_anyscale_cluster,
-    warm_up_cluster,
+    warm_up_one_cluster,
     NUM_CPU_PER_NODE,
     NUM_CONNECTIONS,
 )
@@ -115,7 +115,7 @@ def main(num_replicas: Optional[int], trial_length: Optional[str],
     deploy_replicas(num_replicas, max_batch_size)
 
     logger.info("Warming up cluster ....\n")
-    warm_up_cluster(10, http_host, http_port)
+    warm_up_one_cluster.remote(10, http_host, http_port, "echo")
 
     logger.info(f"Starting wrk trial on all nodes for {trial_length} ....\n")
     # For detailed discussion, see https://github.com/wg/wrk/issues/205