Improve stability of cheets_AppCompatTest.
I have added a retry flow around the core functionality of the test.
Each run will have its own logcat and screenshots.
BUG=None
TEST=Many of the tests locally and in the lab. No failures.
Change-Id: I727d2f2901e644945c3863b6590edf34a9c1bd92
Reviewed-on: https://chromium-review.googlesource.com/c/1401295
Reviewed-by: David Haddock <dhaddock@chromium.org>
Commit-Queue: David Haddock <dhaddock@chromium.org>
Tested-by: David Haddock <dhaddock@chromium.org>
Trybot-Ready: David Haddock <dhaddock@chromium.org>
(cherry picked from commit 17b572600066ec784894ef67b5baa0d1ec851167)
Reviewed-on: https://chromium-review.googlesource.com/c/1427362
diff --git a/client/site_tests/cheets_AppCompatTest/cheets_AppCompatTest.py b/client/site_tests/cheets_AppCompatTest/cheets_AppCompatTest.py
index 800d7a9..0cc3bcc 100644
--- a/client/site_tests/cheets_AppCompatTest/cheets_AppCompatTest.py
+++ b/client/site_tests/cheets_AppCompatTest/cheets_AppCompatTest.py
@@ -4,11 +4,11 @@
import logging
import os
-import time
from autotest_lib.client.bin import utils
from autotest_lib.client.common_lib import error
from autotest_lib.client.common_lib.cros import arc
+from autotest_lib.client.cros.graphics import graphics_utils
class cheets_AppCompatTest(arc.ArcTest):
@@ -124,10 +124,11 @@
arc.adb_shell('rm -r %s' % screenshot, ignore_status=True)
- def _save_logcat(self):
+ def _save_logcat(self, iteration):
"""Saves the logcat for the test run to the resultsdir."""
+ self._logcat = 'logcat%d.txt' % iteration
arc.adb_cmd('logcat -d > %s' % os.path.join(self.resultsdir,
- 'logcat.txt'))
+ self._logcat))
def _parse_results(self):
@@ -144,13 +145,13 @@
self._pkg_name, self._app_version, passed, failed, nt,
blocked, skipped, ft))
logging.info(result)
- if int(failed) > 0:
- raise error.TestFail(result)
+ pass_status = int(failed) == 0 and int(passed) > 0
+ return pass_status, result
def _get_log_entry_count(self, entry):
"""Get the total number of times a string appears in logcat."""
- logcat = os.path.join(self.resultsdir, 'logcat.txt')
+ logcat = os.path.join(self.resultsdir, self._logcat)
return utils.run('grep "%s" %s | grep -c "%s"' %
(self._LOG_TAG, logcat, entry),
ignore_status=True).stdout.strip()
@@ -158,7 +159,7 @@
def _get_failed_test_cases(self):
"""Get the list of test cases that failed from logcat."""
- logcat = os.path.join(self.resultsdir, 'logcat.txt')
+ logcat = os.path.join(self.resultsdir, self._logcat)
failed_tests = []
failed_tests_list = utils.run('grep "%s" %s | grep ",FAIL,"' %
(self._LOG_TAG, logcat),
@@ -173,6 +174,11 @@
arc.adb_cmd('logcat -G 16M')
+ def _clear_logcat_buffer(self):
+ """Clear logcat buffer between runs."""
+ arc.adb_cmd('logcat -c')
+
+
def _get_app_version(self):
"""Grab the version of the application we are testing."""
self._app_version = arc.adb_shell('dumpsys package %s | grep '
@@ -180,13 +186,32 @@
self._pkg_name)
- def run_once(self):
+ def run_once(self, retries=3):
self._increase_logcat_buffer()
self._copy_resources_to_dut()
self._grant_storage_permission()
- self._start_test()
- self._get_app_version()
- self._capture_bugreport()
- self._grab_screenshots()
- self._save_logcat()
- self._parse_results()
+
+ for trial in range(retries):
+ logging.info('Iteration %d: Trying to launch play store' % trial)
+
+ # Bring Play Store to front.
+ arc.adb_shell('am start %s' % self._PLAY_STORE_ACTIVITY)
+ graphics_utils.take_screenshot(self.resultsdir,
+ 'test_start')
+ self._start_test()
+ logging.info('Iteration %d: Test finished' % trial)
+ graphics_utils.take_screenshot(self.resultsdir,
+ 'test_end')
+ self._get_app_version()
+ self._capture_bugreport()
+ self._grab_screenshots()
+ self._save_logcat(trial)
+ passed, result = self._parse_results()
+ if passed:
+ return
+ elif trial + 1 >= retries:
+ raise error.TestFail(result)
+
+ # Kill playstore and clear logcat for next iteration.
+ arc.adb_shell('am force-stop %s' % self._PLAY_STORE_ACTIVITY)
+ self._clear_logcat_buffer()
