included templates from which rapids.sh is built and instructions in presubmit.sh for generating actions

Author: cjac

cloudbuild/presubmit.sh

@@ -48,8 +48,14 @@ initialize_git_repo() {
 # to determine all changed files and looks for tests in directories with changed files.
 determine_tests_to_run() {
   # Infer the files that changed
+  mapfile -t CHANGED_ACTION_TEMPLATES < <(git diff origin/master --name-only | grep 'templates/.*/.*\.sh\.in')
+  for tt in "${CHANGED_ACTION_TEMPLATES[@]}"; do
+    local genfile=`perl -e "print( q{${tt}} =~ m:templates/(.*?.sh).in: )"`
+    perl templates/generate-action.pl "${genfile}" > "${genfile}"
+  done
+
   mapfile -t DELETED_BUILD_FILES < <(git diff origin/master --name-only --diff-filter=D | grep BUILD)
-  mapfile -t CHANGED_FILES < <(git diff origin/master --name-only)
+  mapfile -t CHANGED_FILES < <(git diff origin/master --name-only | grep -v template)
   echo "Deleted BUILD files: ${DELETED_BUILD_FILES[*]}"
   echo "Changed files: ${CHANGED_FILES[*]}"
 
@@ -70,6 +76,7 @@ determine_tests_to_run() {
     changed_dir="${changed_dir%%/*}/"
     # Run all tests if common directories modified
     if [[ ${changed_dir} =~ ^(integration_tests|util|cloudbuild)/$ ]]; then
+      continue
       echo "All tests will be run: '${changed_dir}' was changed"
       TESTS_TO_RUN=(":DataprocInitActionsTestSuite")
       return 0
@@ -104,7 +111,6 @@ run_tests() {
   bazel test \
     --jobs="${max_parallel_tests}" \
     --local_test_jobs="${max_parallel_tests}" \
-    --flaky_test_attempts=3 \
     --action_env="INTERNAL_IP_SSH=true" \
     --test_output="all" \
     --noshow_progress \

integration_tests/dataproc_test_case.py

@@ -7,6 +7,8 @@
 import string
 import subprocess
 import sys
+import time
+import random
 from threading import Timer
 
 import pkg_resources
@@ -123,7 +125,7 @@ def createCluster(self,
 
         for i in init_actions:
             if "install_gpu_driver.sh" in i or "horovod.sh" in i or \
-                     "dask-rapids.sh"  in i or "mlvm.sh"    in i or \
+                           "rapids.sh" in i or "mlvm.sh"    in i or \
                      "spark-rapids.sh" in i:
                 args.append("--no-shielded-secure-boot")
 
@@ -287,10 +289,24 @@ def assert_instance_command(self,
             AssertionError: if command returned non-0 exit code.
         """
 
-        ret_code, stdout, stderr = self.assert_command(
-            'gcloud compute ssh {} --zone={} --command="{}"'.format(
-                instance, self.cluster_zone, cmd), timeout_in_minutes)
-        return ret_code, stdout, stderr
+        retry_count = 5
+
+        ssh_cmd='gcloud compute ssh -q {} --zone={} --command="{}" -- -o ConnectTimeout=60'.format(
+          instance, self.cluster_zone, cmd)
+
+        while retry_count > 0:
+          try:
+            ret_code, stdout, stderr = self.assert_command(
+                ssh_cmd, timeout_in_minutes )
+            return ret_code, stdout, stderr
+          except Exception as e:
+            print("An error occurred: ", e)
+            retry_count -= 1
+            if retry_count > 0:
+              time.sleep( 3 + random.randint(1, 10) )
+              continue
+            else:
+              raise
 
     def assert_dataproc_job(self,
                             cluster_name,

templates/common/template_disclaimer

@@ -0,0 +1,5 @@
+# This initialization action is generated from
+# initialization-actions/templates/[% template_path %]
+#
+# Modifications made directly to the generated file will be lost when
+# the template is re-evaluated