def execute(self, context):
# Check for negative max_total_failures option - negative is an error.
# for per-instance failures, negative means "no limit", so it's allowed.
if context.options.max_total_failures < 0:
context.print_err("max_total_failures option must be >0, but you specified %s" %
context.options.max_total_failures)
return EXIT_INVALID_PARAMETER
job = context.options.instance_spec.jobkey
instances = (None if context.options.instance_spec.instance == ALL_INSTANCES else
context.options.instance_spec.instance)
if instances is not None and context.options.strict:
context.verify_instances_option_validity(job, instances)
api = context.get_api(job.cluster)
config = (context.get_job_config(job, context.options.config)
if context.options.config else None)
updater_config = UpdaterConfig(
context.options.batch_size,
context.options.restart_threshold,
context.options.watch_secs,
context.options.max_per_instance_failures,
context.options.max_total_failures)
resp = api.restart(job, instances, updater_config,
context.options.healthcheck_interval_seconds, config=config)
context.log_response_and_raise(resp,
err_msg="Error restarting job %s:" % str(job))
context.print_out("Job %s restarted successfully" % str(job))
if context.options.open_browser:
webbrowser.open_new_tab(get_job_page(api, job))
return EXIT_OK
def execute(self, context):
# Check for negative max_total_failures option - negative is an error.
# for per-shard failures, negative means "no limit", so it's allowed.
if context.options.max_total_failures < 0:
context.print_err(
"max_total_failures option must be >0, but you specified %s" %
context.options.max_total_failures)
context.print_log(
logging.INFO, "Error: max_total_failures option=%s" %
context.options.max_total_failures)
return EXIT_INVALID_PARAMETER
job = context.options.instance_spec.jobkey
instances = (None
if context.options.instance_spec.instance == ALL_INSTANCES
else context.options.instance_spec.instance)
if instances is not None and context.options.strict:
context.verify_shards_option_validity(job, instances)
api = context.get_api(job.cluster)
config = (context.get_job_config(job, context.options.config)
if context.options.config else None)
updater_config = UpdaterConfig(
context.options.batch_size, context.options.restart_threshold,
context.options.watch_secs,
context.options.max_per_instance_failures,
context.options.max_total_failures,
context.options.rollback_on_failure)
resp = api.restart(job,
instances,
updater_config,
context.options.healthcheck_interval_seconds,
config=config)
if resp.responseCode != ResponseCode.OK:
context.print_err("Error restarting job %s; see log for details" %
str(job))
else:
context.print_out("Job %s restarted successfully" % str(job))
context.check_and_log_response(resp)
if context.options.open_browser:
context.open_job_page(api, context.options.jobspec)
return EXIT_OK
def really_restart(args, options):
if options.max_total_failures < 0:
print("max_total_failures option must be >0, but you specified %s" %
options.max_total_failures,
file=sys.stderr)
exit(1)
maybe_disable_hooks(options)
api, job_key, config_file = LiveJobDisambiguator.disambiguate_args_or_die(
args, options, make_client_factory())
config = get_job_config(job_key.to_path(), config_file,
options) if config_file else None
updater_config = UpdaterConfig(options.batch_size,
options.restart_threshold,
options.watch_secs,
options.max_per_shard_failures,
options.max_total_failures)
resp = api.restart(job_key,
options.shards,
updater_config,
options.health_check_interval_seconds,
config=config)
check_and_log_response(resp)
handle_open(api.scheduler_proxy.scheduler_client().url, job_key.role,
job_key.env, job_key.name)
from apache.aurora.common.aurora_job_key import AuroraJobKey
from gen.apache.aurora.api.AuroraSchedulerManager import Client as scheduler_client
from gen.apache.aurora.api.ttypes import (AssignedTask, Response, ResponseCode,
Result, ScheduledTask,
ScheduleStatus, ScheduleStatusResult,
TaskConfig)
SESSION_KEY = 'test_session'
CLUSTER = 'smfd'
JOB = AuroraJobKey(CLUSTER, 'johndoe', 'test', 'test_job')
HEALTH_CHECK_INTERVAL_SECONDS = 5
UPDATER_CONFIG = UpdaterConfig(
batch_size=2,
restart_threshold=23,
watch_secs=45,
max_per_shard_failures=0,
max_total_failures=0,
rollback_on_failure=True,
)
class TestRestarter(MoxTestBase):
def setUp(self):
super(TestRestarter, self).setUp()
self.mock_scheduler = self.mox.CreateMock(scheduler_client)
self.mock_instance_watcher = self.mox.CreateMock(InstanceWatcher)
self.lock = None
self.restarter = Restarter(
JOB, UPDATER_CONFIG, HEALTH_CHECK_INTERVAL_SECONDS,