Clean up release tests (#11420)

ci/long_running_distributed_tests/ray-project/project.yaml

@@ -1,39 +0,0 @@
-# This file is generated by `ray project create`.
-
-name: long-running-distributed-tests
-
-cluster:
-  config: ray-project/cluster.yaml
-  params:
-    - name: ray_version # Ray version string.
-      default: "0.8.2"
-
-    - name: commit # Ray commit SHA string.
-      default: "f5a1307a608fe5fdbdb04616b22c91f029af329a"
-
-    - name: ray_branch
-      default: "releases/0.8.2"
-
-commands:
-  - name: test_workload
-    help: "Start a long running distributed test."
-    command: |
-      python workloads/{{workload}}.py
-    params:
-      - name: workload
-        help: "Name of workload to run."
-        choices:
-          [
-            "pytorch_pbt_failure"
-          ]
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: [
-  # For example, uncomment this to save the logs from the
-  # last ray job.
-  # "/tmp/ray/session_latest",
-]

ci/long_running_distributed_tests/run.sh

@@ -1,45 +0,0 @@
-#!/usr/bin/env bash
-
-ray_version="" 
-commit=""
-ray_branch=""
-workload=""
-
-usage() {
-    echo "Start one microbenchmark trial."
-}
-
-for i in "$@"
-do
-echo "$i"
-case "$i" in
-    --ray-version=*)
-    ray_version="${i#*=}"
-
-    ;;
-    --commit=*)
-    commit="${i#*=}"
-    ;;
-    --ray-branch=*)
-    ray_branch="${i#*=}"
-    ;;
-    --workload=*)
-    workload="${i#*=}"
-    ;;
-    --help)
-    usage
-    exit
-    ;;
-    *)
-    echo "unknown arg, $i"
-    exit 1
-    ;;
-esac
-done
-
-echo "version: $ray_version"
-echo "commit: $commit"
-echo "branch: $ray_branch"
-echo "workload: $workload"
-
-python "workloads/$workload.py"

ci/long_running_tests/ray-project/project.yaml

@@ -1,54 +0,0 @@
-name: long-running-tests
-description: "Ray's long running stress tests"
-
-cluster:
-  config: ray-project/cluster.yaml
-
-commands:
-  - name: test_workload
-    help: "Start a long running stress test."
-    command: |
-      # Install nightly Ray wheels.
-      source activate tensorflow_p36 && pip install -q -U {{wheel}} Click
-      source activate tensorflow_p36 && pip install -q ray[all] gym[atari]
-      source activate tensorflow_p36 && python workloads/{{workload}}.py
-    params:
-      - name: wheel
-        help: "URL to the ray wheel to test (defaults to latest)."
-        default: https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-1.1.0.dev0-cp36-cp36m-manylinux1_x86_64.whl
-      - name: workload
-        help: "Name of the workload to run."
-        choices:
-          [
-            "actor_deaths",
-            "apex",
-            "impala",
-            "many_actor_tasks",
-            "many_drivers",
-            "many_tasks",
-            "node_failures",
-            "pbt",
-            "serve",
-            "serve_failure",
-            "many_tasks_serialized_ids",
-          ]
-    config:
-      tmux: true
-
-  - name: check-load
-    command: uptime
-    help: "Check load of the workload."
-
-  - name: show-output
-    command: tmux capture-pane -p
-    help: "Show tail of the workoad output."
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: []
-# For example, uncomment this to save the logs from the
-# last ray job.
-# "/tmp/ray/session_latest",

ci/microbenchmark/ray-project/project.yaml

@@ -1,39 +0,0 @@
-name: microbenchmark
-description: "Ray's microbenchmark"
-
-cluster:
-  config: ray-project/cluster.yaml
-
-commands:
-  - name: run
-    help: "Start one microbenchmark trial."
-    command: |      
-      rm ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl || true
-      wget https://s3-us-west-2.amazonaws.com/ray-wheels/{{ray_branch}}/{{commit}}/ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-      
-      pip uninstall -y -q ray
-      pip install -U ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-      
-      OMP_NUM_THREADS=64 ray microbenchmark
-    params:
-    - name: ray_version # Ray version string.
-      default: "1.1.0.dev0"
-
-    - name: commit # Ray commit SHA string.
-      default: "FILL ME IN"
-
-    - name: ray_branch
-      default: "master"
-    config:
-      tmux: true
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: [
-  # For example, uncomment this to save the logs from the
-  # last ray job.
-  # "/tmp/ray/session_latest",
-]

ci/regression_test/stress_tests/ray-project/project.yaml

@@ -1,37 +0,0 @@
-# This file is generated by `ray project create`.
-
-name: ray_stress_tests
-
-cluster:
-  config: ray-project/cluster.yaml
-  params:
-    - name: ray_version # Ray version string.
-      default: "0.8.2"
-
-    - name: commit # Ray commit SHA string.
-      default: "f5a1307a608fe5fdbdb04616b22c91f029af329a"
-
-    - name: ray_branch
-      default: "releases/0.8.2"
-
-commands:
-  - name: test_many_tasks
-    help: "Run a stress test that executes many tasks."
-    command: |
-      python test_many_tasks.py
-
-  - name: test_dead_actors
-    help: "Run a stress test that spawns and kills many actors."
-    command: |
-      python test_dead_actors.py
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: [
-  # For example, uncomment this to save the logs from the
-  # last ray job.
-  # "/tmp/ray/session_latest",
-]

ci/regression_test/stress_tests/ray-project/requirements.txt

@@ -1 +0,0 @@
-ray[debug]

ci/regression_test/stress_tests/run.sh

@@ -1,45 +0,0 @@
-#!/usr/bin/env bash
-
-ray_version="" 
-commit=""
-ray_branch=""
-workload=""
-
-usage() {
-    echo "Start one microbenchmark trial."
-}
-
-for i in "$@"
-do
-echo "$i"
-case "$i" in
-    --ray-version=*)
-    ray_version="${i#*=}"
-
-    ;;
-    --commit=*)
-    commit="${i#*=}"
-    ;;
-    --ray-branch=*)
-    ray_branch="${i#*=}"
-    ;;
-    --workload=*)
-    workload="${i#*=}"
-    ;;
-    --help)
-    usage
-    exit
-    ;;
-    *)
-    echo "unknown arg, $i"
-    exit 1
-    ;;
-esac
-done
-
-echo "version: $ray_version"
-echo "commit: $commit"
-echo "branch: $ray_branch"
-echo "workload: $workload"
-
-python "$workload.py"

ci/rllib_tests/regresssion_tests/ray-project/project.yaml

@@ -1,57 +0,0 @@
-# This file is generated by `ray project create`.
-
-name: rllib_regression_tests
-
-# description: A short description of the project.
-# The URL of the repo this project is part of.
-# repo: ...
-
-cluster:
-  config: ray-project/cluster.yaml
-  params:
-    - name: ray_version # Ray version string.
-      default: "0.8.2"
-
-    - name: commit # Ray commit SHA string.
-      default: "f5a1307a608fe5fdbdb04616b22c91f029af329a"
-
-    - name: ray_branch
-      default: "releases/0.8.2"
-
-
-environment:
-  # dockerfile: The dockerfile to be built and ran the commands with.
-  # dockerimage: The docker image to be used to run the project in, e.g. ubuntu:18.04.
-  requirements: ray-project/requirements.txt
-
-  shell: # Shell commands to be ran for environment setup.
-    - echo "Setting up the environment"
-
-commands:
-  - name: check-load
-    command: uptime
-    help: "Check load of the workload."
-
-  - name: check-gpu
-    command: nvidia-smi
-    help: "Check load of the gpu."
-
-  - name: show-output
-    command: tmux capture-pane -p
-    help: "Show tail of the workoad output."
-
-  - name: run-regression-tests
-    command: |
-      source activate tensorflow_p36 && rllib train -f compact-regression-test.yaml
-    help: "Run rllib regression tests"
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: [
-  # For example, uncomment this to save the logs from the
-  # last ray job.
-  # "/tmp/ray/session_latest",
-]

ci/rllib_tests/regresssion_tests/run.sh

@@ -1,3 +0,0 @@
-#!/usr/bin/env bash
-
-source activate tensorflow_p36 && rllib train -f compact-regression-test.yaml

ci/rllib_tests/stress_tests/ray-project/project.yaml

@@ -1,52 +0,0 @@
-# This file is generated by `ray project create`.
-
-name: rllib_stress_tests
-
-# description: A short description of the project.
-# The URL of the repo this project is part of.
-# repo: ...
-
-cluster:
-  config: ray-project/cluster.yaml
-  params:
-    - name: ray_version # Ray version string.
-      default: "0.8.2"
-
-    - name: commit # Ray commit SHA string.
-      default: "f5a1307a608fe5fdbdb04616b22c91f029af329a"
-
-    - name: ray_branch
-      default: "releases/0.8.2"
-
-
-environment:
-  # dockerfile: The dockerfile to be built and ran the commands with.
-  # dockerimage: The docker image to be used to run the project in, e.g. ubuntu:18.04.
-  requirements: ray-project/requirements.txt
-
-  shell: # Shell commands to be ran for environment setup.
-    - echo "Setting up the environment"
-
-commands:
-  - name: check-load
-    command: uptime
-    help: "Check load of the workload."
-
-  - name: show-output
-    command: tmux capture-pane -p
-    help: "Show tail of the workoad output."
-
-  - name: run-impala
-    command: bash run.sh
-    help: "Run impala stress test"
-
-# Pathnames for files and directories that should be saved
-# in a snapshot but that should not be synced with a# session. Pathnames can be relative to the project
-# directory or absolute. Generally, this should be files
-# that were created by an active session, such as
-# application checkpoints and logs.
-output_files: [
-  # For example, uncomment this to save the logs from the
-  # last ray job.
-  # "/tmp/ray/session_latest",
-]

ci/rllib_tests/stress_tests/run.sh

@@ -1,7 +0,0 @@
-#!/usr/bin/env bash
-
-source activate tensorflow_p36
-
-python3 wait_cluster.py
-
-rllib train -f atari_impala_xlarge.yaml --ray-address=auto --queue-trials

ci/rllib_tests/unit_gpu_tests/run.sh

@@ -1,5 +0,0 @@
-#!/usr/bin/env bash
-
-# Run all test cases, but with a forced num_gpus=1.
-# TODO: (sven) chose correct dir and run over all RLlib tests and example scripts!
-source activate tensorflow_p36 && export RAY_FORCE_NUM_GPUS=1 && cd ~ && python -m pytest test_attention_net_learning.py

release/1.0.0.env

@@ -0,0 +1,2 @@
+export ray_version="1.0.0rc1"
+export commit=fd5ddb661e659c2b0c968661d96d0405426912e5

release/README.md

@@ -0,0 +1,2 @@
+Running the kickoff script:
+

release/RELEASE_PROCESS.rst

@@ -0,0 +1 @@
+../doc/dev/RELEASE_PROCESS.rst

release/kickoff.sh

@@ -0,0 +1,29 @@
+#!/usr/bin/env bash
+
+
+source "$2"
+
+ray_version=${ray_version:-}
+commit=${commit:-}
+
+if [[ $ray_version == "" || $commit == "" || $1 == "" ]]
+then
+    echo "Provide --ray-version, --commit, and --ray-branch"
+    exit 1
+fi
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "workload: $1"
+
+DATESTR=$(date +%Y%m%d-%H%M)
+SESSION="$1-$DATESTR"
+
+echo "session: $SESSION"
+
+chmod +x ./run.sh
+if [ -z "$NO_UP" ]; then
+  anyscale up "$SESSION"
+fi
+anyscale push "$SESSION"
+anyscale exec -n "$SESSION" "./run.sh $1 --ray-version=$ray_version --commit=$commit"

release/long_running_distributed_tests/cluster.yaml

@@ -58,23 +58,7 @@ worker_nodes:
   InstanceMarketOptions:
     MarketType: spot
 
-setup_commands:
-  # Install ray.
-  - conda uninstall -y terminado || true
-  - pip install -U pip
-  - pip install terminado
-  - wget --quiet https://s3-us-west-2.amazonaws.com/ray-wheels/{{ray_branch}}/{{commit}}/ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-  - ray || pip install -U ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl || true
-  # Installing this without -U to make sure we don't replace the existing Ray installation
-  - pip install ray[rllib]
-  - pip install -U ipdb
-  # There have been some recent problems with torch 1.5 and torchvision 0.6
-  # not recognizing GPUs.
-  # So, we force install torch 1.4 and torchvision 0.5. 
-  # https://github.com/pytorch/pytorch/issues/37212#issuecomment-623198624.
-  - pip install torch==1.4.0 torchvision==0.5.0
-  - echo set-window-option -g mouse on > ~/.tmux.conf
-  - echo 'termcapinfo xterm* ti@:te@' > ~/.screenrc
+setup_commands: []
 
 # Command to start ray on the head node. You don't need to change this.
 head_start_ray_commands:

release/long_running_distributed_tests/run.sh

@@ -0,0 +1,61 @@
+#!/usr/bin/env bash
+
+ray_version="" 
+commit=""
+ray_branch=""
+workload=""
+
+usage() {
+    echo "Start one microbenchmark trial."
+}
+
+for i in "$@"
+do
+echo "$i"
+case "$i" in
+    --ray-version=*)
+    ray_version="${i#*=}"
+
+    ;;
+    --commit=*)
+    commit="${i#*=}"
+    ;;
+    --ray-branch=*)
+    ray_branch="${i#*=}"
+    ;;
+    --workload=*)
+    workload="${i#*=}"
+    ;;
+    --help)
+    usage
+    exit
+    ;;
+    *)
+    echo "unknown arg, $i"
+    exit 1
+    ;;
+esac
+done
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "branch: $ray_branch"
+echo "workload: $workload"
+
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
+
+conda uninstall -y terminado || true
+pip install -U pip
+pip install terminado
+pip install -U "$wheel"
+pip install "ray[rllib]"
+pip install -U ipdb
+# There have been some recent problems with torch 1.5 and torchvision 0.6
+# not recognizing GPUs.
+# So, we force install torch 1.4 and torchvision 0.5. 
+# https://github.com/pytorch/pytorch/issues/37212#issuecomment-623198624.
+pip install torch==1.4.0 torchvision==0.5.0
+echo set-window-option -g mouse on > ~/.tmux.conf
+echo 'termcapinfo xterm* ti@:te@' > ~/.screenrc
+
+python "workloads/$workload.py"

release/long_running_tests/cluster.yaml

@@ -33,22 +33,7 @@ worker_nodes:
         MarketType: spot
 
 # List of shell commands to run to set up nodes.
-setup_commands:
-    # Install latest TensorFlow
-    - source activate tensorflow_p36 && conda remove -y --force wrapt || true
-    - source activate tensorflow_p36 && pip install --upgrade pip
-    - source activate tensorflow_p36 && pip install -U tensorflow==1.14
-    - echo set-window-option -g mouse on > ~/.tmux.conf
-    - echo 'termcapinfo xterm* ti@:te@' > ~/.screenrc
-    # Serve load testing tool
-    - 'rm -r wrk || true && git clone https://github.com/wg/wrk.git wrk && cd wrk && make -j && sudo cp wrk /usr/local/bin'
-
-    # Uncomment the following if you wish to build Ray instead.
-    # - sudo apt-get update
-    # - sudo apt-get install -y build-essential curl unzip
-    # - git clone https://github.com/ray-project/ray || true
-    # - ray/ci/travis/install-bazel.sh
-    # - cd ray/python; git checkout master; git pull; source activate tensorflow_p36 && pip install -e . --verbose
+setup_commands: []
 
 # Custom commands that will be run on the head node after common setup.
 head_setup_commands: []

release/long_running_tests/run.sh

@@ -50,9 +50,17 @@ echo "workload: $workload"
 
 wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
 
+echo set-window-option -g mouse on > ~/.tmux.conf
+echo 'termcapinfo xterm* ti@:te@' > ~/.screenrc
+# Serve load testing tool
+rm -r wrk || true && git clone https://github.com/wg/wrk.git wrk && cd wrk && make -j && sudo cp wrk /usr/local/bin
 pip install -U pip
 unset RAY_ADDRESS
-source activate tensorflow_p36 && pip install -q -U "$wheel" Click
-source activate tensorflow_p36 && pip install -q "ray[all]" "gym[atari]"
-source activate tensorflow_p36 && python "workloads/$workload.py"
+source activate tensorflow_p36 
+conda remove -y --force wrapt || true
+pip install --upgrade pip
+pip install -U tensorflow==1.14
+pip install -q -U "$wheel" Click
+pip install -q "ray[all]" "gym[atari]"
+python "workloads/$workload.py"
 

release/microbenchmark/run.sh

@@ -21,6 +21,9 @@ case "$i" in
     --ray-branch=*)
     ray_branch="${i#*=}"
     ;;
+    --workload=*)
+    workload="${i#*=}"
+    ;;
     --help)
     usage
     exit
@@ -38,15 +41,20 @@ then
     exit 1
 fi
 
+
 echo "version: $ray_version"
 echo "commit: $commit"
 echo "branch: $ray_branch"
+echo "workload: $workload"
 
-rm "ray-$ray_version-cp38-cp38-manylinux1_x86_64.whl" || true
-wget "https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp38-cp38-manylinux1_x86_64.whl"
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp38-cp38-manylinux1_x86_64.whl"
 
+
+echo set-window-option -g mouse on > ~/.tmux.conf
+echo 'termcapinfo xterm* ti@:te@' > ~/.screenrc
 pip uninstall -y -q ray
-pip install -U "ray-$ray_version-cp38-cp38-manylinux1_x86_64.whl"
+pip install --upgrade pip
+pip install -U "$wheel"
 
 unset RAY_ADDRESS
 OMP_NUM_THREADS=64 ray microbenchmark

release/rllib_tests/regression_tests/cluster.yaml

@@ -26,13 +26,7 @@ head_node:
 
 
 # List of shell commands to run to set up nodes.
-setup_commands:
-    - wget --quiet https://s3-us-west-2.amazonaws.com/ray-wheels/{{ray_branch}}/{{commit}}/ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-    - conda uninstall -y terminado
-    - source activate tensorflow_p36 && pip install -U ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-    - source activate tensorflow_p36 && pip install ray[rllib] ray[debug]
-    - source activate tensorflow_p36 && pip install torch==1.6 torchvision
-    - source activate tensorflow_p36 && pip install boto3==1.4.8 cython==0.29.0
+setup_commands: []
 
 # Command to start ray on the head node. You don't need to change this.
 head_start_ray_commands:

release/rllib_tests/regression_tests/run.sh

@@ -0,0 +1,52 @@
+#!/usr/bin/env bash
+ray_version="" 
+commit=""
+ray_branch=""
+
+for i in "$@"
+do
+echo "$i"
+case "$i" in
+    --ray-version=*)
+    ray_version="${i#*=}"
+
+    ;;
+    --commit=*)
+    commit="${i#*=}"
+    ;;
+    --ray-branch=*)
+    ray_branch="${i#*=}"
+    ;;
+    --workload=*)
+    ;;
+    --help)
+    usage
+    exit
+    ;;
+    *)
+    echo "unknown arg, $i"
+    exit 1
+    ;;
+esac
+done
+
+if [[ $ray_version == "" || $commit == "" || $ray_branch == "" ]]
+then
+    echo "Provide --ray-version, --commit, and --ray-branch"
+    exit 1
+fi
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "branch: $ray_branch"
+echo "workload: ignored"
+
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
+
+conda uninstall -y terminado
+source activate tensorflow_p36 && pip install -U pip
+source activate tensorflow_p36 && pip install -U "$wheel"
+source activate tensorflow_p36 && pip install "ray[rllib]" "ray[debug]"
+source activate tensorflow_p36 && pip install torch==1.6 torchvision
+source activate tensorflow_p36 && pip install boto3==1.4.8 cython==0.29.0
+source activate tensorflow_p36 && rllib train -f compact-regression-test.yaml

release/rllib_tests/stress_tests/cluster.yaml

@@ -86,12 +86,7 @@ file_mounts: {
 }
 
 # List of shell commands to run to set up nodes.
-setup_commands:
-    - wget --quiet https://s3-us-west-2.amazonaws.com/ray-wheels/{{ray_branch}}/{{commit}}/ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-    - conda uninstall -y terminado
-    - source activate tensorflow_p36 && pip install -U ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-    - source activate tensorflow_p36 && pip install ray[rllib] ray[debug]
-    - source activate tensorflow_p36 && pip install boto3==1.4.8 cython==0.29.0
+setup_commands: []
 
 # Custom commands that will be run on the head node after common setup.
 head_setup_commands: []

release/rllib_tests/stress_tests/run.sh

@@ -0,0 +1,56 @@
+#!/usr/bin/env bash
+
+ray_version="" 
+commit=""
+ray_branch=""
+
+for i in "$@"
+do
+echo "$i"
+case "$i" in
+    --ray-version=*)
+    ray_version="${i#*=}"
+
+    ;;
+    --commit=*)
+    commit="${i#*=}"
+    ;;
+    --ray-branch=*)
+    ray_branch="${i#*=}"
+    ;;
+    --workload=*)
+    ;;
+    --help)
+    usage
+    exit
+    ;;
+    *)
+    echo "unknown arg, $i"
+    exit 1
+    ;;
+esac
+done
+
+if [[ $ray_version == "" || $commit == "" || $ray_branch == "" ]]
+then
+    echo "Provide --ray-version, --commit, and --ray-branch"
+    exit 1
+fi
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "branch: $ray_branch"
+echo "workload: ignored"
+
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
+
+conda uninstall -y terminado
+source activate tensorflow_p36 && pip install -U pip
+source activate tensorflow_p36 && pip install -U "$wheel"
+source activate tensorflow_p36 && pip install "ray[rllib]" "ray[debug]"
+source activate tensorflow_p36 && pip install boto3==1.4.8 cython==0.29.0
+source activate tensorflow_p36
+
+python3 wait_cluster.py
+
+rllib train -f atari_impala_xlarge.yaml --ray-address=auto --queue-trials

release/rllib_tests/unit_gpu_tests/cluster.yaml

@@ -26,10 +26,7 @@ head_node:
 
 
 # List of shell commands to run to set up nodes.
-setup_commands:
-    - wget --quiet https://s3-us-west-2.amazonaws.com/ray-wheels/{{ray_branch}}/{{commit}}/ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
-    - conda uninstall -y terminado
-    - source activate tensorflow_p36 && pip install -U ray-{{ray_version}}-cp36-cp36m-manylinux1_x86_64.whl
+setup_commands: []
 
 # Command to start ray on the head node. You don't need to change this.
 head_start_ray_commands:

release/rllib_tests/unit_gpu_tests/run.sh

@@ -0,0 +1,53 @@
+#!/usr/bin/env bash
+
+ray_version="" 
+commit=""
+ray_branch=""
+
+for i in "$@"
+do
+echo "$i"
+case "$i" in
+    --ray-version=*)
+    ray_version="${i#*=}"
+
+    ;;
+    --commit=*)
+    commit="${i#*=}"
+    ;;
+    --ray-branch=*)
+    ray_branch="${i#*=}"
+    ;;
+    --workload=*)
+    ;;
+    --help)
+    usage
+    exit
+    ;;
+    *)
+    echo "unknown arg, $i"
+    exit 1
+    ;;
+esac
+done
+
+if [[ $ray_version == "" || $commit == "" || $ray_branch == "" ]]
+then
+    echo "Provide --ray-version, --commit, and --ray-branch"
+    exit 1
+fi
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "branch: $ray_branch"
+echo "workload: ignored"
+
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
+
+conda uninstall -y terminado
+source activate tensorflow_p36 && pip install -U pip
+source activate tensorflow_p36 && pip install -U "$wheel"
+
+# Run all test cases, but with a forced num_gpus=1.
+# TODO: (sven) chose correct dir and run over all RLlib tests and example scripts!
+source activate tensorflow_p36 && export RAY_FORCE_NUM_GPUS=1 && cd ~ && python -m pytest test_attention_net_learning.py

release/stress_tests/cluster.yaml

@@ -85,7 +85,7 @@ file_mounts: {
 }
 
 # List of shell commands to run to set up nodes.
-setup_commands:
+setup_commands: []
     # Uncomment these if you want to build ray from source.
     # - sudo apt-get -qq update
     # - sudo apt-get install -y build-essential curl unzip

release/stress_tests/run.sh

@@ -0,0 +1,57 @@
+#!/usr/bin/env bash
+
+ray_version="" 
+commit=""
+ray_branch=""
+workload=""
+
+for i in "$@"
+do
+echo "$i"
+case "$i" in
+    --ray-version=*)
+    ray_version="${i#*=}"
+
+    ;;
+    --commit=*)
+    commit="${i#*=}"
+    ;;
+    --ray-branch=*)
+    ray_branch="${i#*=}"
+    ;;
+    --workload=*)
+    workload="${i#*=}"
+    ;;
+    --help)
+    usage
+    exit
+    ;;
+    *)
+    echo "unknown arg, $i"
+    exit 1
+    ;;
+esac
+done
+
+echo "version: $ray_version"
+echo "commit: $commit"
+echo "branch: $ray_branch"
+echo "workload: $workload"
+
+wheel="https://s3-us-west-2.amazonaws.com/ray-wheels/$ray_branch/$commit/ray-$ray_version-cp36-cp36m-manylinux1_x86_64.whl"
+
+# Install Anaconda.
+wget --quiet https://repo.continuum.io/archive/Anaconda3-5.0.1-Linux-x86_64.sh || true
+bash Anaconda3-5.0.1-Linux-x86_64.sh -b -p "$HOME/anaconda3" || true
+# shellcheck disable=SC2016
+echo 'export PATH="$HOME/anaconda3/bin:$PATH"' >> ~/.bashrc
+
+conda uninstall -y terminado
+source activate tensorflow_p36 && pip install -U pip
+source activate tensorflow_p36 && pip install -U "$wheel"
+
+pip install -U pip 
+conda uninstall -y terminado || true
+pip install terminado
+pip install boto3==1.4.8 cython==0.29.0
+python "workloads/$workload.py"