From b750bd7fc91a79ae88f70e422470e920376522e2 Mon Sep 17 00:00:00 2001 From: Edward Oakes Date: Wed, 15 Jan 2020 19:47:59 -0600 Subject: [PATCH] Use 2xlarge instances in long running tests (#6802) --- ci/long_running_tests/ray-project/cluster.yaml | 2 +- ci/long_running_tests/ray-project/project.yaml | 3 +-- ci/long_running_tests/workloads/actor_deaths.py | 3 ++- ci/long_running_tests/workloads/apex.py | 3 ++- ci/long_running_tests/workloads/impala.py | 3 ++- ci/long_running_tests/workloads/many_actor_tasks.py | 3 ++- ci/long_running_tests/workloads/many_drivers.py | 3 ++- ci/long_running_tests/workloads/many_tasks.py | 3 ++- ci/long_running_tests/workloads/node_failures.py | 3 ++- ci/long_running_tests/workloads/pbt.py | 3 ++- 10 files changed, 18 insertions(+), 11 deletions(-) diff --git a/ci/long_running_tests/ray-project/cluster.yaml b/ci/long_running_tests/ray-project/cluster.yaml index 95a70d209..10f9c6db6 100644 --- a/ci/long_running_tests/ray-project/cluster.yaml +++ b/ci/long_running_tests/ray-project/cluster.yaml @@ -13,7 +13,7 @@ auth: ssh_user: ubuntu head_node: - InstanceType: m5.xlarge + InstanceType: m5.2xlarge ImageId: ami-0888a3b5189309429 # DLAMI 7/1/19 BlockDeviceMappings: - DeviceName: /dev/sda1 diff --git a/ci/long_running_tests/ray-project/project.yaml b/ci/long_running_tests/ray-project/project.yaml index 96a0c56ef..db934cf16 100644 --- a/ci/long_running_tests/ray-project/project.yaml +++ b/ci/long_running_tests/ray-project/project.yaml @@ -10,8 +10,7 @@ commands: command: | # Install nightly Ray wheels. source activate tensorflow_p36 && pip install -U {{wheel}} - source activate tensorflow_p36 && pip install ray[rllib] ray[debug] gym[atari] - source activate tensorflow_p36 && pip install ray[debug] + source activate tensorflow_p36 && pip install ray[dashboard,debug,rllib,tune] gym[atari] source activate tensorflow_p36 && python workloads/{{workload}}.py params: - name: wheel diff --git a/ci/long_running_tests/workloads/actor_deaths.py b/ci/long_running_tests/workloads/actor_deaths.py index e05489706..f3fd71949 100644 --- a/ci/long_running_tests/workloads/actor_deaths.py +++ b/ci/long_running_tests/workloads/actor_deaths.py @@ -28,7 +28,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/apex.py b/ci/long_running_tests/workloads/apex.py index 7d33b89fb..e7ecd9b02 100644 --- a/ci/long_running_tests/workloads/apex.py +++ b/ci/long_running_tests/workloads/apex.py @@ -25,7 +25,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/impala.py b/ci/long_running_tests/workloads/impala.py index 1e54fbc19..de566f560 100644 --- a/ci/long_running_tests/workloads/impala.py +++ b/ci/long_running_tests/workloads/impala.py @@ -25,7 +25,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/many_actor_tasks.py b/ci/long_running_tests/workloads/many_actor_tasks.py index a2875d8ee..08035b0f3 100644 --- a/ci/long_running_tests/workloads/many_actor_tasks.py +++ b/ci/long_running_tests/workloads/many_actor_tasks.py @@ -28,7 +28,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/many_drivers.py b/ci/long_running_tests/workloads/many_drivers.py index 9ceb558dc..8154186cb 100644 --- a/ci/long_running_tests/workloads/many_drivers.py +++ b/ci/long_running_tests/workloads/many_drivers.py @@ -27,7 +27,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 5}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/many_tasks.py b/ci/long_running_tests/workloads/many_tasks.py index ca2c36cba..3daba67ea 100644 --- a/ci/long_running_tests/workloads/many_tasks.py +++ b/ci/long_running_tests/workloads/many_tasks.py @@ -28,7 +28,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/node_failures.py b/ci/long_running_tests/workloads/node_failures.py index 87863229e..7c123baac 100644 --- a/ci/long_running_tests/workloads/node_failures.py +++ b/ci/long_running_tests/workloads/node_failures.py @@ -26,7 +26,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload. diff --git a/ci/long_running_tests/workloads/pbt.py b/ci/long_running_tests/workloads/pbt.py index 69b87f997..8808a6a39 100644 --- a/ci/long_running_tests/workloads/pbt.py +++ b/ci/long_running_tests/workloads/pbt.py @@ -26,7 +26,8 @@ for i in range(num_nodes): num_gpus=0, resources={str(i): 2}, object_store_memory=object_store_memory, - redis_max_memory=redis_max_memory) + redis_max_memory=redis_max_memory, + webui_host="0.0.0.0") ray.init(address=cluster.address) # Run the workload.