[SPARK-451] Use run-tests.py to run tests (apache#140)

* User run-tests.py to run tests

* Add .cache to .gitignore

* remove comments

* fix test jar file path

* provide HDFS_SERVICE_NAME when uninstalling

* remove install_cli from publish jobs

.gitignore

@@ -1,3 +1,4 @@
+.cache/
 build/
 dcos-commons-tools/
 tests/env

bin/jenkins-dist-publish.sh

@@ -34,8 +34,6 @@ SPARK_VERSION=${GIT_BRANCH#origin/tags/custom-} # e.g. "2.0.2"
 source "${DIR}/jenkins.sh"
 
 pushd "${SPARK_BUILD_DIR}"
-install_cli
 docker_login
-# build_and_test
 publish_dists
 popd

bin/jenkins-dist-test.sh

@@ -8,7 +8,6 @@ SPARK_DIR="${DIR}/../../spark"
 
 function run() {
     source bin/jenkins.sh
-    install_cli
     docker_login
     build_and_test
 }

bin/jenkins-package-publish.sh

@@ -36,7 +36,6 @@ source "${DIR}/jenkins.sh"
 
 pushd "${SPARK_BUILD_DIR}"
 SPARK_VERSION=$(jq -r ".spark_version" manifest.json)
-install_cli
 docker_login
 publish_docker_images
 make_universe

bin/jenkins-package-test.sh

@@ -7,9 +7,7 @@ SPARK_BUILD_DIR=${DIR}/..
 
 function run() {
     source bin/jenkins.sh
-    install_cli
     docker_login
-
     make --directory=dispatcher universe
     export $(cat "${WORKSPACE}/stub-universe.properties")
     make test

bin/jenkins.sh

@@ -112,20 +112,6 @@ function docker_version() {
     echo "${SPARK_BUILD_VERSION}-hadoop-$1"
 }
 
-function install_cli {
-    curl -O https://downloads.mesosphere.io/dcos-cli/install.sh
-    rm -rf dcos-cli/
-    mkdir dcos-cli
-    bash install.sh dcos-cli http://change.me --add-path no
-    source dcos-cli/bin/env-setup
-
-    # hack because the installer forces an old CLI version
-    pip install -U dcoscli
-
-    # needed in `make test`
-    pip3 install jsonschema
-}
-
 function docker_login {
     docker login [email protected] --username="${DOCKER_USERNAME}" --password="${DOCKER_PASSWORD}"
 }

bin/test.sh

@@ -7,6 +7,7 @@ set -x
 set -o pipefail
 
 BIN_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+SPARK_BUILD_DIR="${BIN_DIR}/.."
 
 check_env() {
     # Check env early, before starting the cluster:
@@ -36,16 +37,6 @@ start_cluster() {
     fi
 }
 
-configure_cli() {
-    dcos config set core.dcos_url "${DCOS_URL}"
-    dcos config set core.ssl_verify false
-    ${COMMONS_DIR}/tools/dcos_login.py
-    dcos config show
-    if [ -n "${STUB_UNIVERSE_URL}" ]; then
-        dcos package repo add --index=0 spark-test "${STUB_UNIVERSE_URL}"
-    fi
-}
-
 initialize_service_account() {
     if [ "$SECURITY" = "strict" ]; then
         ${COMMONS_DIR}/tools/create_service_account.sh --strict
@@ -59,21 +50,15 @@ build_scala_test_jar() {
 }
 
 run_tests() {
-    pushd tests
-    if [[ ! -d venv ]]; then
-        virtualenv -p python3 venv
-    fi
-    source venv/bin/activate
-    pip install -r requirements.txt
-    SCALA_TEST_JAR_PATH=$(pwd)/jobs/scala/target/scala-2.11/dcos-spark-scala-tests-assembly-0.1-SNAPSHOT.jar \
-                       py.test -s test.py
-    popd
+    SCALA_TEST_JAR_PATH=${SPARK_BUILD_DIR}/tests/jobs/scala/target/scala-2.11/dcos-spark-scala-tests-assembly-0.1-SNAPSHOT.jar \
+                       CLUSTER_URL=${DCOS_URL} \
+                       STUB_UNIVERSE_URL=${STUB_UNIVERSE_URL} \
+                       ${COMMONS_DIR}/tools/run_tests.py shakedown ${SPARK_BUILD_DIR}/tests ${SPARK_BUILD_DIR}/tests/requirements.txt
 }
 
 check_env
 start_cluster
 # TODO: Migrate the following three commands to dcos-commons-tools/run-tests.py
-configure_cli
 initialize_service_account
 build_scala_test_jar
 run_tests

tests/test.py → tests/test_spark.py

@@ -25,11 +25,16 @@
 def _init_logging():
     logging.basicConfig(level=logging.INFO)
     logging.getLogger('dcos').setLevel(logging.WARNING)
+    logging.getLogger('requests').setLevel(logging.WARNING)
 
 
 _init_logging()
 LOGGER = logging.getLogger(__name__)
 THIS_DIR = os.path.dirname(os.path.abspath(__file__))
+DEFAULT_HDFS_TASK_COUNT=10
+HDFS_PACKAGE_NAME='beta-hdfs'
+HDFS_SERVICE_NAME='hdfs'
+SPARK_PACKAGE_NAME='spark'
 
 
 def setup_module(module):
@@ -39,12 +44,13 @@ def setup_module(module):
 
 
 def teardown_module(module):
-    shakedown.uninstall_package_and_wait('spark')
+    shakedown.uninstall_package_and_wait(SPARK_PACKAGE_NAME)
     if _hdfs_enabled():
-        shakedown.uninstall_package_and_wait('hdfs')
-        _run_janitor('hdfs')
+        shakedown.uninstall_package_and_wait(HDFS_PACKAGE_NAME, HDFS_SERVICE_NAME)
+        _run_janitor(HDFS_PACKAGE_NAME)
 
 
+@pytest.mark.sanity
 def test_jar():
     master_url = ("https" if _is_strict() else "http") + "://leader.mesos:5050"
     spark_job_runner_args = '{} dcos \\"*\\" spark:only 2 --auth-token={}'.format(
@@ -57,6 +63,7 @@ def test_jar():
                ["--class", 'com.typesafe.spark.test.mesos.framework.runners.SparkJobRunner'])
 
 
+@pytest.mark.sanity
 def test_teragen():
     if _hdfs_enabled():
         jar_url = "https://downloads.mesosphere.io/spark/examples/spark-terasort-1.0-jar-with-dependencies_2.11.jar"
@@ -66,6 +73,7 @@ def test_teragen():
                    ["--class", "com.github.ehiggs.spark.terasort.TeraGen"])
 
 
+@pytest.mark.sanity
 def test_python():
     python_script_path = os.path.join(THIS_DIR, 'jobs', 'python', 'pi_with_include.py')
     python_script_url = _upload_file(python_script_path)
@@ -100,6 +108,7 @@ def test_kerberos():
          "--conf", "sun.security.krb5.debug=true"])
 
 
+@pytest.mark.sanity
 def test_r():
     r_script_path = os.path.join(THIS_DIR, 'jobs', 'R', 'dataframe.R')
     r_script_url = _upload_file(r_script_path)
@@ -108,6 +117,7 @@ def test_r():
                "Justin")
 
 
+@pytest.mark.sanity
 def test_cni():
     SPARK_EXAMPLES="http://downloads.mesosphere.com/spark/assets/spark-examples_2.11-2.0.1.jar"
     _run_tests(SPARK_EXAMPLES,
@@ -117,6 +127,7 @@ def test_cni():
                 "--class", "org.apache.spark.examples.SparkPi"])
 
 
+@pytest.mark.sanity
 def test_s3():
     linecount_path = os.path.join(THIS_DIR, 'resources', 'linecount.txt')
     s3.upload_file(linecount_path)
@@ -147,14 +158,14 @@ def _hdfs_enabled():
 def _require_hdfs():
     LOGGER.info("Ensuring HDFS is installed.")
 
-    _require_package('hdfs', _get_hdfs_options())
+    _require_package(HDFS_PACKAGE_NAME, _get_hdfs_options())
     _wait_for_hdfs()
 
 
 def _require_spark():
     LOGGER.info("Ensuring Spark is installed.")
 
-    _require_package('spark', _get_spark_options())
+    _require_package(SPARK_PACKAGE_NAME, _get_spark_options())
     _wait_for_spark()
 
 
@@ -187,17 +198,18 @@ def _get_hdfs_options():
     if _is_strict():
         options = {'service': {'principal': 'service-acct', 'secret_name': 'secret'}}
     else:
-        options = {}
+        options = {"service": {}}
+
+    options["service"]["beta-optin"] = True
     return options
 
 
 def _wait_for_hdfs():
     shakedown.wait_for(_is_hdfs_ready, ignore_exceptions=False, timeout_seconds=900)
 
 
-DEFAULT_HDFS_TASK_COUNT=10
 def _is_hdfs_ready(expected_tasks = DEFAULT_HDFS_TASK_COUNT):
-    running_tasks = [t for t in shakedown.get_service_tasks('hdfs') \
+    running_tasks = [t for t in shakedown.get_service_tasks(HDFS_SERVICE_NAME) \
                      if t['state'] == 'TASK_RUNNING']
     return len(running_tasks) >= expected_tasks
 
@@ -231,7 +243,10 @@ def _install_spark():
                               "principal": "service-acct"}
         options['security'] = {"mesos": {"authentication": {"secret_name": "secret"}}}
 
-    shakedown.install_package('spark', options_json=options, wait_for_completion=True)
+    shakedown.install_package(
+        SPARK_PACKAGE_NAME,
+        options_json=options,
+        wait_for_completion=True)
 
     def pred():
         dcos_url = dcos.config.get_config_val("core.dcos_url")