def v2_runner_on_failed(self, result, ignore_errors=False):
# schedule_logger(self._job_id).error(f"Run on failed, result: {result._task._uuid}")
# schedule_logger(self._job_id).error(f"Run on failed, details: {result.__dict__}")
# stat_logger.info(f'>>>>>>>>>run on ok failed: {json.dumps(result._result, indent=4)}')
schedule_logger(self._job_id).error(
f"<TASK FAILED> Host {result._host} executes task {result._task._uuid} {result._task._attributes.get('name')} failed. Details: {json.dumps(result._result, indent=4)}"
)
update_info = {
'job_id': self._job_id,
'play_id': self._play_id,
'task_id': result._task._uuid,
'end_time': current_timestamp(),
'status': TaskStatus.FAILED,
}
JobSaver.update_task(update_info)
JobSaver.update_task_status(update_info)
JobSaver.update_play(update_info)
JobSaver.update_play_status(update_info)
JobSaver.update_job(update_info)
JobSaver.update_job_status(update_info)
delegated_vars = result._result.get('_ansible_delegated_vars', None)
self._clean_results(result._result, result._task.action)
if self._last_task_banner != result._task._uuid:
self._print_task_banner(result._task)
self._handle_exception(result._result,
use_stderr=self.display_failed_stderr)
self._handle_warnings(result._result)
if result._task.loop and 'results' in result._result:
self._process_items(result)
else:
if delegated_vars:
self._display.display(
"fatal: [%s -> %s]: FAILED! => %s" %
(result._host.get_name(), delegated_vars['ansible_host'],
self._dump_results(result._result)),
color=C.COLOR_ERROR,
stderr=self.display_failed_stderr)
else:
self._display.display("fatal: [%s]: FAILED! => %s" %
(result._host.get_name(),
self._dump_results(result._result)),
color=C.COLOR_ERROR,
stderr=self.display_failed_stderr)
if ignore_errors:
self._display.display("...ignoring", color=C.COLOR_SKIP)
def v2_playbook_on_no_hosts_matched(self):
schedule_logger(
self._job_id).info("<No host matched> Skipping: no hosts matched.")
update_info = {
'job_id': self._job_id,
'play_id': self._play_id,
'end_time': current_timestamp(),
'status': PlayStatus.FAILED
}
JobSaver.update_play(update_info)
JobSaver.update_play_status(update_info)
JobSaver.update_job(update_info)
JobSaver.update_job_status(update_info)
self._display.display("skipping: no hosts matched", color=C.COLOR_SKIP)
def stop_play(job_id, play_id, status=PlayStatus.CANCELED):
plays = JobSaver.query_play(play_id=play_id)
if plays:
play = plays[0]
kill_status = job_utils.kill_play_process_execution(play)
if kill_status:
if OngoingStatus.contains(play.f_status):
play_info = {
'job_id': job_id,
'play_id': play_id,
'end_time': current_timestamp(),
'status': status,
}
JobSaver.update_play_status(play_info)
if not StandbyStatus.contains(play.f_status):
JobSaver.update_play(play_info)
return True
else:
return False
else:
schedule_logger(job_id).info(f"cannot find and kill process of play {play_id}")
return False
def v2_runner_on_unreachable(self, result):
try:
schedule_logger(self._job_id).error(
f"<HOST UNREACHABLE> Host {result._host} executes task {result._task._uuid} {result._task._attributes.get('name')} unreachable. \nDetails: {json.dumps(result._result, indent=4)}"
)
# schedule_logger(self._job_id).info(f'>>>>>>>>>in runner on unreachable: {json.dumps(result._result, indent=4)}')
# stat_logger.info(f'>>>>>>>>>in runner on unreachable: {json.dumps(result._result, indent=4)}')
update_info = {
'job_id': self._job_id,
'play_id': self._play_id,
'task_id': result._task._uuid,
'end_time': current_timestamp(),
'status': TaskStatus.FAILED,
}
JobSaver.update_task(update_info)
JobSaver.update_task_status(update_info)
JobSaver.update_play(update_info)
JobSaver.update_play_status(update_info)
except Exception:
schedule_logger(self._job_id).warning(
f"In v2_runner_on_unreachable, details: {traceback.format_exc()}"
)
if self._last_task_banner != result._task._uuid:
self._print_task_banner(result._task)
delegated_vars = result._result.get('_ansible_delegated_vars', None)
if delegated_vars:
msg = "fatal: [%s -> %s]: UNREACHABLE! => %s" % (
result._host.get_name(), delegated_vars['ansible_host'],
self._dump_results(result._result))
else:
msg = "fatal: [%s]: UNREACHABLE! => %s" % (
result._host.get_name(), self._dump_results(result._result))
self._display.display(msg,
color=C.COLOR_UNREACHABLE,
stderr=self.display_failed_stderr)
def retry_play(job_id, play_id, test_mode=False):
plays = JobSaver.query_play(play_id=play_id)
if not plays:
return 100, f"Retry play {play_id} failed, can not find such play in database."
# copy play conf into package dir
play_conf_path_dict = file_utils.get_play_conf_path(play_id)
with open(file_utils.get_job_conf_path(job_id), 'r') as f:
job_conf = json.loads(f.read())
package_dir = get_package_dir_by_version(job_conf.get('version'))
play_conf_path_dict['conf_path'] = shutil.copy2(
src=play_conf_path_dict['conf_path'], dst=package_dir)
update_info = {
'job_id': job_id,
'play_id': play_id,
'status': PlayStatus.WAITING,
}
JobSaver.update_play_status(update_info)
JobSaver.update_job_status(update_info)
# clean task records
JobSaver.clean_task(play_id=play_id)
# execute run_play method
try:
play_retry_executor_pool.submit(
PlayController.run_play,
job_id=job_id,
play_id=play_id,
play_conf_path=play_conf_path_dict['conf_path'],
play_hosts_path=play_conf_path_dict['hosts_path'],
test_mode=test_mode,
retry_mode=True)
return 0, f"Start retrying play {play_id}"
except Exception as e:
stat_logger.exception(e)
return 100, f"Retry play {play_id} failed, details: {str(e)}"
def run_job(job_id):
job_data = job_utils.get_job_configuration(job_id=job_id)
stat_logger.info(
f"in play controller run job func, get job data: {json.dumps(job_data, indent=4)}"
)
schedule_logger(job_id).info(
f"in play controller, func run job: {json.dumps(job_data, indent=4)}"
)
play_conf_path_dict = PlayController.initialize_plays(
job_id=job_id, job_data=job_data)
stat_logger.info(
f"in play controller run job func after initialize play\n get play conf path dict: {play_conf_path_dict}"
)
# TODO get package dir by version
package_dir = get_package_dir_by_version(job_data.get('version'))
if not os.path.exists(package_dir) and not os.path.isdir(package_dir):
raise Exception(
f'Local package directory {package_dir} not exists.')
job_info = {
'job_id': job_id,
'status': JobStatus.RUNNING,
'start_time': current_timestamp()
}
JobSaver.update_job_status(job_info)
JobSaver.update_job(job_info)
for play_id, conf_dict in play_conf_path_dict.items():
conf_dict['conf_path'] = shutil.copy2(src=conf_dict['conf_path'],
dst=package_dir)
PlayController.run_play(
job_id=job_id,
play_id=play_id,
play_conf_path=conf_dict.get('conf_path'),
play_hosts_path=conf_dict.get('hosts_path'),
test_mode=TEST_MODE)
if os.path.exists(conf_dict['conf_path']):
os.remove(conf_dict['conf_path'])
plays = JobSaver.query_play(play_id=play_id)
if plays:
play = plays[0]
status = play.f_status
if status != PlayStatus.SUCCESS:
if status in [
PlayStatus.CANCELED, PlayStatus.FAILED,
PlayStatus.TIMEOUT
]:
update_info = {
'job_id': job_id,
'play_id': play_id,
'status': status,
'end_time': current_timestamp()
}
JobSaver.update_play_status(update_info)
JobSaver.update_play(update_info)
JobSaver.update_job_status(update_info)
JobSaver.update_job(update_info)
else:
update_info = {
'job_id': job_id,
'play_id': play_id,
'status': PlayStatus.FAILED,
'end_time': current_timestamp()
}
schedule_logger(job_id).error(
f'Unexpected error occured on play {play_id}, job {job_id} failed, previous status of play: {play.f_status}'
)
stat_logger.error(
f'Unexpected error occured on play {play_id}, job {job_id} failed, previous status of play: {play.f_status}'
)
JobSaver.update_play_status(update_info)
JobSaver.update_play(update_info)
JobSaver.update_job_status(update_info)
JobSaver.update_job(update_info)
schedule_logger(job_id).info(
f"job {job_id} finished, status is {update_info.get('status')}"
)
break
else:
update_info = {
'job_id': job_id,
'play_id': play_id,
'status': PlayStatus.SUCCESS,
'end_time': current_timestamp()
}
JobSaver.update_play_status(update_info)
JobSaver.update_play(update_info)
else:
raise Exception(f'can not find play {play_id}')
else:
update_info = {
'job_id': job_id,
'status': JobStatus.SUCCESS,
'end_time': current_timestamp()
}
JobSaver.update_job(update_info)
JobSaver.update_job_status(update_info)
schedule_logger(job_id).info(
f"job {job_id} finished, status is {update_info.get('status')}"
)
if not TEST_MODE:
plays = JobSaver.query_play(job_id=job_id,
status=PlayStatus.SUCCESS)
modules = []
module_names = []
for play in plays:
module_name = play.f_roles.strip('[]').replace('_', '')
module_names.append(module_name)
modules.append({
'name':
module_name,
'ips':
job_data.get('modules', {}).get(module_name,
{}).get('ips', []),
'port':
job_data.get('modules', {}).get(module_name,
{}).get('port', None)
})
# parties = PartyInfo.get_or_none(f_version=job_data.get('version'), f_party_id=job_data.get('party_id'))
parties = PartyInfo.get_or_none(
f_party_id=job_data.get('party_id'))
if parties:
module_mapping = dict(zip(module_names, modules))
stored_modules = parties.f_modules.get("data", [])
name_map = {}
for offset, item in enumerate(stored_modules):
name_map[item.get('name')] = offset
for key, value in module_mapping.items():
if key in name_map:
schedule_logger(job_id).info(
f"{key} in name map, in replace process")
stored_modules[name_map[key]] = value
else:
schedule_logger(job_id).info(
f"{key} not in name map, in append process ")
stored_modules.append(value)
# update_status = False
# for offset, module_info in enumerate(stored_modules):
# if module_info['name'] in module_mapping:
# stored_modules[offset] = module_mapping[module_info['name']]
# update_status = True
for key in ['role', 'version']:
# if parties[key] != job_data[key]:
# parties[key] = job_data[key]
if getattr(parties, f'f_{key}') != job_data[key]:
setattr(parties, f'f_{key}', job_data[key])
# update_status = True
# if update_status:
parties.f_modules = {'data': stored_modules}
parties.save()
DB.commit()
else:
party_info = PartyInfo()
# party_info.f_job_id = job_id
party_info.f_role = job_data.get('role')
party_info.f_version = job_data.get('version')
party_info.f_party_id = job_data.get('party_id')
party_info.f_modules = {'data': modules}
party_info.save(force_insert=True)
def run_play(job_id,
play_id,
play_conf_path,
play_hosts_path,
test_mode=False,
retry_mode=False):
schedule_logger(job_id).info(
f'Trying to start to run play with id: {play_id}')
# task_process_start_status = False
process_cmd = [
'python3',
sys.modules[PlayExecutor.__module__].__file__,
'--job_id',
job_id,
'--play_id',
play_id,
'--conf_path',
play_conf_path,
'--hosts_path',
play_hosts_path,
]
if test_mode:
process_cmd.append('--test')
if retry_mode:
process_cmd.append('--retry')
schedule_logger(job_id).info(
f"Trying to start job {job_id}, play {play_id} subprocess.")
try:
config_dir = file_utils.get_play_directory(play_id)
std_dir = file_utils.get_job_log_directory(job_id)
p = job_utils.run_subprocess(config_dir=config_dir,
process_cmd=process_cmd,
log_dir=std_dir)
if p:
# task_process_start_status = True
play_info = {
'pid': p.pid,
'job_id': job_id,
'play_id': play_id,
'status': PlayStatus.RUNNING,
'start_time': current_timestamp()
}
JobSaver.update_play_status(play_info=play_info)
JobSaver.update_play(play_info=play_info)
p.wait()
else:
raise Exception(f'play {play_id} start subprocess failed')
except Exception as e:
play_info = {
'job_id': job_id,
'play_id': play_id,
'status': PlayStatus.FAILED,
'end_time': current_timestamp()
}
JobSaver.update_play_status(play_info)
JobSaver.update_play(play_info)
schedule_logger(job_id).exception(e)
finally:
if retry_mode:
job_info = {
'job_id':
job_id,
'play_id':
play_id,
'end_time':
current_timestamp(),
'status':
JobStatus.SUCCESS if PlayController.check_job_status(
job_id=job_id) else JobStatus.FAILED
}
JobSaver.update_job(job_info)
JobSaver.update_job_status(job_info)
def v2_playbook_on_stats(self, stats):
self._display.banner("PLAY RECAP")
hosts = sorted(stats.processed.keys())
failed_count = 0
for h in hosts:
t = stats.summarize(h)
# stat_logger.info(f'>>>>>>>>PLAY RECAP, {t}')
schedule_logger(self._job_id).info(
f"in playbook on stats: RECAP: {t}, host: {h}")
failed_count += t['failures']
failed_count += t['unreachable']
# self._display.display(
# u"%s : %s %s %s %s %s %s %s" % (
# hostcolor(h, t),
# colorize(u'ok', t['ok'], C.COLOR_OK),
# colorize(u'changed', t['changed'], C.COLOR_CHANGED),
# colorize(u'unreachable', t['unreachable'], C.COLOR_UNREACHABLE),
# colorize(u'failed', t['failures'], C.COLOR_ERROR),
# colorize(u'skipped', t['skipped'], C.COLOR_SKIP),
# colorize(u'rescued', t['rescued'], C.COLOR_OK),
# colorize(u'ignored', t['ignored'], C.COLOR_WARN),
# ),
# screen_only=True
# )
self._display.display(u"%s : %s %s %s %s %s %s %s" % (
hostcolor(h, t, False),
colorize(u'ok', t['ok'], None),
colorize(u'changed', t['changed'], None),
colorize(u'unreachable', t['unreachable'], None),
colorize(u'failed', t['failures'], None),
colorize(u'skipped', t['skipped'], None),
colorize(u'rescued', t['rescued'], None),
colorize(u'ignored', t['ignored'], None),
),
log_only=True)
if not failed_count:
update_info = {
'job_id': self._job_id,
'play_id': self._play_id,
'end_time': current_timestamp(),
'status': PlayStatus.SUCCESS
}
JobSaver.update_play(update_info)
JobSaver.update_play_status(update_info)
self._display.display("", screen_only=True)
# print custom stats if required
if stats.custom and self.show_custom_stats:
self._display.banner("CUSTOM STATS: ")
# per host
# TODO: come up with 'pretty format'
for k in sorted(stats.custom.keys()):
if k == '_run':
continue
self._display.display(
'\t%s: %s' %
(k, self._dump_results(stats.custom[k], indent=1).replace(
'\n', '')))
# print per run custom stats
if '_run' in stats.custom:
self._display.display("", screen_only=True)
self._display.display('\tRUN: %s' % self._dump_results(
stats.custom['_run'], indent=1).replace('\n', ''))
self._display.display("", screen_only=True)
if context.CLIARGS['check'] and self.check_mode_markers:
self._display.banner("DRY RUN")