[Android] Redesigns the sharder to allow replicated vs distributed tests

build/android/pylib/base/test_dispatcher.py

 # Copyright (c) 2013 The Chromium Authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 """Implements test sharding logic."""

[frankf, 2013/07/16 18:42:32]

Update

[gkanwar, 2013/07/16 20:27:36]

Done.

 
 import logging
 import threading
 
 from pylib import android_commands
 from pylib import constants
 from pylib import forwarder
 from pylib.utils import reraiser_thread
 from pylib.utils import watchdog_timer
 
 import base_test_result
 
 
 DEFAULT_TIMEOUT = 7 * 60  # seven minutes
+VALID_TEST_ALLOCATION = ['shard', 'replicate']
 
 
 class _ThreadSafeCounter(object):
   """A threadsafe counter."""
 
   def __init__(self):
     self._lock = threading.Lock()
     self._value = 0
 
   def GetAndIncrement(self):
@@ skipping 81 matching lines @@
   def __iter__(self):
     """Iterate through tests in the collection until all have been handled."""
     while True:
       r = self._pop()
       if r is None:
         break
       yield r
 
 
 def _RunTestsFromQueue(runner, test_collection, out_results, watcher,
-                       num_retries):
+                       num_retries, tag_results_with_device=False):
   """Runs tests from the test_collection until empty using the given runner.
 
   Adds TestRunResults objects to the out_results list and may add tests to the
   out_retry list.
 
   Args:
     runner: A TestRunner object used to run the tests.
     test_collection: A _TestCollection from which to get _Test objects to run.
     out_results: A list to add TestRunResults to.
     watcher: A watchdog_timer.WatchdogTimer object, used as a shared timeout.
     num_retries: Number of retries for a test.
+    tag_results_with_device: If True, appends the name of the device on which
+        the test was run to the test name. Used by ReplicateAndRunTests to
+        identify which device ran each copy of the test, and to ensure each copy
+        of the test is recorded separately.
   """
+
+  # Used to tag all results to identify which device caused failing tests
+  def TagTestRunResults(test_run_results):
+    new_test_run_results = base_test_result.TestRunResults()
+    for test_result in test_run_results.GetAll():
+      new_result = base_test_result.BaseTestResult(
+          '%s_%s' % (runner.device, test_result.GetName()),

[frankf, 2013/07/16 18:42:32]

Use last 4 digits of device serial.

[gkanwar, 2013/07/16 20:27:36]

Done.

+          test_result.GetType(), test_result.GetLog())

[frankf, 2013/07/16 18:42:32]

As the name suggest, BaseTestResult can be derived from. This doesn't work.

[gkanwar, 2013/07/16 20:27:36]

Updated to move tag into the BaseTestResult class itself.

+      new_test_run_results.AddResult(new_result)
+    return new_test_run_results
+
   for test in test_collection:
     watcher.Reset()
     try:
       if not android_commands.IsDeviceAttached(runner.device):
         # Device is unresponsive, stop handling tests on this device.
         msg = 'Device %s is unresponsive.' % runner.device
         logging.warning(msg)
         raise android_commands.errors.DeviceUnresponsiveError(msg)
       result, retry = runner.RunTest(test.test)
+      if tag_results_with_device:
+        result = TagTestRunResults(result)
       test.tries += 1
       if retry and test.tries <= num_retries:
         # Retry non-passing results, only record passing results.
         pass_results = base_test_result.TestRunResults()
+        # Tag all results with the device, so we can identify the failing device
+        # for replicated tests.

[frankf, 2013/07/16 18:42:32]

Is this comment misplaced?

[gkanwar, 2013/07/16 20:27:36]

Oops, removed.

         pass_results.AddResults(result.GetPass())
         out_results.append(pass_results)
         logging.warning('Will retry test, try #%s.' % test.tries)
         test_collection.add(_Test(test=retry, tries=test.tries))
       else:
         # All tests passed or retry limit reached. Either way, record results.
         out_results.append(result)
     except:
       # An unhandleable exception, ensure tests get run by another device and
       # reraise this exception on the main thread.
@@ skipping 20 matching lines @@
   try:
     index = threadsafe_counter.GetAndIncrement()
     logging.warning('Creating shard %s for device %s.', index, device)
     runner = runner_factory(device, index)
     runner.SetUp()
     out_runners.append(runner)
   except android_commands.errors.DeviceUnresponsiveError as e:
     logging.warning('Failed to create shard for %s: [%s]', device, e)
 
 
-def _RunAllTests(runners, tests, num_retries, timeout=None):
+def _RunAllTests(runners, test_collection_factory, num_retries, timeout=None,
+                 tag_results_with_device=False):
   """Run all tests using the given TestRunners.
 
   Args:
     runners: a list of TestRunner objects.
-    tests: a list of Tests to run using the given TestRunners.
+    test_collection_factory: a callable to generate a _TestCollection object for
+        each test runner.
     num_retries: number of retries for a test.
     timeout: watchdog timeout in seconds, defaults to the default timeout.
+    tag_results_with_device: If True, appends the name of the device on which
+        the test was run to the test name. Used by ReplicateAndRunTests to
+        identify which device ran each copy of the test, and to ensure each copy
+        of the test is recorded separately.
 
   Returns:
     A tuple of (TestRunResults object, exit code)
   """
-  logging.warning('Running %s tests with %s test runners.' %
-                  (len(tests), len(runners)))
-  tests_collection = _TestCollection([_Test(t) for t in tests])
+  logging.warning('Running tests with %s test runners.' % (len(runners)))
   results = []
   exit_code = 0
   watcher = watchdog_timer.WatchdogTimer(timeout)
+
   workers = reraiser_thread.ReraiserThreadGroup(
       [reraiser_thread.ReraiserThread(
           _RunTestsFromQueue,
-          [r, tests_collection, results, watcher, num_retries],
+          [r, test_collection_factory(), results, watcher, num_retries,
+           tag_results_with_device],
           name=r.device[-4:])
        for r in runners])
   run_results = base_test_result.TestRunResults()
   workers.StartAll()
 
   # Catch DeviceUnresponsiveErrors and set a warning exit code
   try:
     workers.JoinAll(watcher)
   except android_commands.errors.DeviceUnresponsiveError as e:
     logging.error(e)
@@ skipping 41 matching lines @@
     runners: a list of TestRunner objects.
     timeout: watchdog timeout in seconds, defaults to the default timeout.
   """
   threads = reraiser_thread.ReraiserThreadGroup(
       [reraiser_thread.ReraiserThread(r.TearDown, name=r.device[-4:])
        for r in runners])
   threads.StartAll()
   threads.JoinAll(watchdog_timer.WatchdogTimer(timeout))
 
 
-def ShardAndRunTests(runner_factory, devices, tests, build_type='Debug',
-                     test_timeout=DEFAULT_TIMEOUT,
-                     setup_timeout=DEFAULT_TIMEOUT,
-                     num_retries=2):
+
+def _GetAttachedDevices(wait_for_debugger=False, test_device=None):
+  """Get all attached devices.
+
+  If we are using a debugger, limit to only one device.
+
+  Args:
+    wait_for_debugger: True if this run will use a debugger.
+    test_device: name of a specific device to use.
+
+  Returns:
+    A list of attached devices.
+  """
+  attached_devices = []
+
+  attached_devices = android_commands.GetAttachedDevices()
+  if test_device:
+    assert test_device in attached_devices

[frankf, 2013/07/16 18:42:32]

add a message to assert

[gkanwar, 2013/07/16 20:27:36]

Done.

+    attached_devices = [test_device]
+
+  if len(attached_devices) > 1 and wait_for_debugger:
+    logging.warning('Debugger can not be sharded, using first available device')
+    attached_devices = attached_devices[:1]
+
+  return attached_devices
+
+
+def RunTests(tests, runner_factory,
+             wait_for_debugger, test_device,
+             test_allocation='shard',

[frankf, 2013/07/16 18:42:32]

I think a boolean called 'shard' makes more sense, and cuts down the validation
below.

[gkanwar, 2013/07/16 20:27:36]

Done.

+             build_type='Debug',
+             test_timeout=DEFAULT_TIMEOUT,
+             setup_timeout=DEFAULT_TIMEOUT,
+             num_retries=2):
   """Run all tests on attached devices, retrying tests that don't pass.
 
   Args:
+    tests: list of tests to run.

[frankf, 2013/07/16 18:42:32]

Capital first letter. 

[gkanwar, 2013/07/16 20:27:36]

Done.

     runner_factory: callable that takes a device and index and returns a
-      TestRunner object.
-    devices: list of attached device serial numbers as strings.
-    tests: list of tests to run.
+        TestRunner object.
+    wait_for_debugger: True if this test is using a debugger.
+    test_device: A specific device to run tests on, or None.
+    test_allocation: 'shard' or 'replicate'.
     build_type: either 'Debug' or 'Release'.
     test_timeout: watchdog timeout in seconds for running tests, defaults to the
-      default timeout.
+        default timeout.
     setup_timeout: watchdog timeout in seconds for creating and cleaning up
-      test runners, defaults to the default timeout.
+        test runners, defaults to the default timeout.
     num_retries: number of retries for a test.
+    tag_results_with_device: If True, appends the name of the device on which
+        the test was run to the test name. Used by ReplicateAndRunTests to
+        identify which device ran each copy of the test, and to ensure each copy

[frankf, 2013/07/16 18:42:32]

Update this

[gkanwar, 2013/07/16 20:27:36]

Done.

+        of the test is recorded separately.
 
   Returns:
     A tuple of (base_test_result.TestRunResults object, exit code).
   """
+  # Validation

[frankf, 2013/07/16 18:42:32]

this is obvious. remove comment.

[gkanwar, 2013/07/16 20:27:36]

Done.

   if not tests:
     logging.error('No tests to run.')
     return (base_test_result.TestRunResults(), constants.ERROR_EXIT_CODE)
 
+  if not test_allocation in VALID_TEST_ALLOCATION:
+    logging.error('Unknown test allocation string %s. Options are: %s'
+                  % (test_allocation, ', '.join(VALID_TEST_ALLOCATION)))
+    return (base_test_result.TestRunResults(), constants.ERROR_EXIT_CODE)
+
+  if test_allocation == 'shard':
+    # Generate a shared _TestCollection object for all test runners, so they
+    # draw from a common pool of tests.
+    shared_test_collection = _TestCollection([_Test(t) for t in tests])
+    test_collection_factory = lambda: shared_test_collection
+    tag_results_with_device = False
+  else:
+    # Generate a unique _TestCollection object for each test runner, but use
+    # the same set of tests.
+    test_collection_factory = lambda: _TestCollection([_Test(t) for t in tests])
+    tag_results_with_device = True
+
+  devices = _GetAttachedDevices(wait_for_debugger, test_device)
+
   logging.info('Will run %d tests: %s', len(tests), str(tests))
+
   forwarder.Forwarder.KillHost(build_type)
   runners = _CreateRunners(runner_factory, devices, setup_timeout)
   try:
-    return _RunAllTests(runners, tests, num_retries, test_timeout)
+    return _RunAllTests(runners, test_collection_factory,
+                        num_retries, test_timeout, tag_results_with_device)
   finally:
     try:
       _TearDownRunners(runners, setup_timeout)
     except android_commands.errors.DeviceUnresponsiveError as e:
       logging.warning('Device unresponsive during TearDown: [%s]', e)
     finally:
       forwarder.Forwarder.KillHost(build_type)