def main(args):
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
# get those jobs listed in configs
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(soa_dir=args.soa_dir)
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
sensu_output, sensu_status = sensu_message_status_for_jobs(service, instance, job_state_pairs)
monitoring_overrides = compose_monitoring_overrides_for_service(
cluster=system_paasta_config.get_cluster(),
service=service,
instance=instance,
soa_dir=args.soa_dir
)
send_event_to_sensu(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=args.soa_dir,
)
def main(args):
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
# get those jobs listed in configs
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(
soa_dir=args.soa_dir)
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
sensu_output, sensu_status = sensu_message_status_for_jobs(
service, instance, job_state_pairs)
monitoring_overrides = compose_monitoring_overrides_for_service(
cluster=system_paasta_config.get_cluster(),
service=service,
instance=instance,
soa_dir=args.soa_dir)
send_event_to_sensu(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=args.soa_dir,
)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster, soa_dir=soa_dir)
try:
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, chronos_job in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
try:
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
except utils.NoDeploymentsAvailable:
log.info("Skipping %s because no deployments are available" % service)
continue
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
chronos_job=chronos_job,
client=client,
)
if sensu_status is not None:
send_event(chronos_job_config, sensu_status, sensu_output)
except (chronos.ChronosAPIError) as e:
log.error("CRITICAL: Unable to contact Chronos! Error: %s" % e)
sys.exit(2)
def main():
args = parse_args()
jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster=args.cluster, soa_dir=args.soa_dir)
# TODO use compose_job_id instead of constructing string once INTERNAL_SPACER deprecated
composed = ['%s%s%s' % (name, chronos_tools.INTERNAL_SPACER, job) for name, job in jobs]
print '\n'.join(composed)
sys.exit(0)
def main():
args = parse_args()
jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster=args.cluster, soa_dir=args.soa_dir)
# TODO use compose_job_id instead of constructing string once INTERNAL_SPACER deprecated
composed = ['%s%s%s' % (name, chronos_tools.INTERNAL_SPACER, job) for name, job in jobs]
paasta_print('\n'.join(composed))
sys.exit(0)
def main():
args = parse_args()
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
# get_chronos_jobs_for_cluster returns (service, job)
expected_service_jobs = chronos_tools.get_chronos_jobs_for_cluster(soa_dir=args.soa_dir)
# filter jobs not related to paasta
# and decompose into (service, instance, tag) tuples
paasta_jobs = filter_paasta_jobs(deployed_job_names(client))
running_service_jobs = [chronos_tools.decompose_job_id(job) for job in paasta_jobs]
to_delete = jobs_to_delete(expected_service_jobs, running_service_jobs)
# recompose the job ids again for deletion
to_delete_job_ids = [chronos_tools.compose_job_id(*job) for job in to_delete]
task_responses = cleanup_tasks(client, to_delete_job_ids)
task_successes = []
task_failures = []
for response in task_responses:
if isinstance(response[-1], Exception):
task_failures.append(response)
else:
task_successes.append(response)
job_responses = cleanup_jobs(client, to_delete_job_ids)
job_successes = []
job_failures = []
for response in job_responses:
if isinstance(response[-1], Exception):
job_failures.append(response)
else:
job_successes.append(response)
if len(to_delete) == 0:
print 'No Chronos Jobs to remove'
else:
if len(task_successes) > 0:
print format_list_output("Successfully Removed Tasks (if any were running) for:",
[job[0] for job in task_successes])
# if there are any failures, print and exit appropriately
if len(task_failures) > 0:
print format_list_output("Failed to Delete Tasks for:", [job[0] for job in task_failures])
if len(job_successes) > 0:
print format_list_output("Successfully Removed Jobs:", [job[0] for job in job_successes])
# if there are any failures, print and exit appropriately
if len(job_failures) > 0:
print format_list_output("Failed to Delete Jobs:", [job[0] for job in job_failures])
if len(job_failures) > 0 or len(task_failures) > 0:
sys.exit(1)
def main():
args = parse_args()
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
# get_chronos_jobs_for_cluster returns (service, job)
expected_service_jobs = chronos_tools.get_chronos_jobs_for_cluster(soa_dir=args.soa_dir)
# filter jobs not related to paasta
# and decompose into (service, instance, tag) tuples
paasta_jobs = filter_paasta_jobs(deployed_job_names(client))
running_service_jobs = [chronos_tools.decompose_job_id(job) for job in paasta_jobs]
to_delete = jobs_to_delete(expected_service_jobs, running_service_jobs)
# recompose the job ids again for deletion
to_delete_job_ids = [chronos_tools.compose_job_id(*job) for job in to_delete]
task_responses = cleanup_tasks(client, to_delete_job_ids)
task_successes = []
task_failures = []
for response in task_responses:
if isinstance(response[-1], Exception):
task_failures.append(response)
else:
task_successes.append(response)
job_responses = cleanup_jobs(client, to_delete_job_ids)
job_successes = []
job_failures = []
for response in job_responses:
if isinstance(response[-1], Exception):
job_failures.append(response)
else:
job_successes.append(response)
if len(to_delete) == 0:
print 'No Chronos Jobs to remove'
else:
if len(task_successes) > 0:
print format_list_output("Successfully Removed Tasks (if any were running) for:",
[job[0] for job in task_successes])
# if there are any failures, print and exit appropriately
if len(task_failures) > 0:
print format_list_output("Failed to Delete Tasks for:", [job[0] for job in task_failures])
if len(job_successes) > 0:
print format_list_output("Successfully Removed Jobs:", [job[0] for job in job_successes])
# if there are any failures, print and exit appropriately
if len(job_failures) > 0:
print format_list_output("Failed to Delete Jobs:", [job[0] for job in job_failures])
if len(job_failures) > 0 or len(task_failures) > 0:
sys.exit(1)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(
cluster, soa_dir=soa_dir)
try:
service_job_mapping = build_service_job_mapping(
client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
try:
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
except utils.NoDeploymentsAvailable:
paasta_print(
utils.PaastaColors.cyan(
"Skipping %s because no deployments are available" %
service))
continue
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
service=service,
instance=instance,
cluster=cluster,
job_state_pairs=job_state_pairs)
if sensu_status is not None:
monitoring_overrides = compose_monitoring_overrides_for_service(
chronos_job_config=chronos_job_config, soa_dir=soa_dir)
send_event(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=soa_dir,
)
except (chronos.ChronosAPIError) as e:
paasta_print(
utils.PaastaColors.red(
"CRITICAL: Unable to contact Chronos! Error: %s" % e))
sys.exit(2)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster, soa_dir=soa_dir)
try:
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
try:
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
except utils.NoDeploymentsAvailable:
paasta_print(utils.PaastaColors.cyan("Skipping %s because no deployments are available" % service))
continue
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
service=service,
instance=instance,
cluster=cluster,
job_state_pairs=job_state_pairs
)
if sensu_status is not None:
monitoring_overrides = compose_monitoring_overrides_for_service(
chronos_job_config=chronos_job_config,
soa_dir=soa_dir
)
send_event(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=soa_dir,
)
except (chronos.ChronosAPIError) as e:
paasta_print(utils.PaastaColors.red("CRITICAL: Unable to contact Chronos! Error: %s" % e))
sys.exit(2)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster, soa_dir=soa_dir)
try:
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
service=service,
instance=instance,
cluster=cluster,
job_state_pairs=job_state_pairs
)
monitoring_overrides = compose_monitoring_overrides_for_service(
chronos_job_config=chronos_job_config,
soa_dir=soa_dir
)
send_event(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=soa_dir,
)
except (ServerNotFoundError, chronos.ChronosAPIError, socket_error) as e:
print(utils.PaastaColors.red("CRITICAL: Unable to contact Chronos! Error: %s" % e))
sys.exit(2)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(cluster, soa_dir=soa_dir)
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
service=service,
instance=instance,
cluster=cluster,
job_state_pairs=job_state_pairs
)
monitoring_overrides = compose_monitoring_overrides_for_service(
chronos_job_config=chronos_job_config,
soa_dir=soa_dir
)
send_event(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=soa_dir,
)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
configured_jobs = chronos_tools.get_chronos_jobs_for_cluster(
cluster, soa_dir=soa_dir)
service_job_mapping = build_service_job_mapping(client, configured_jobs)
for service_instance, job_state_pairs in service_job_mapping.items():
service, instance = service_instance[0], service_instance[1]
chronos_job_config = load_chronos_job_config(
service=service,
instance=instance,
cluster=cluster,
soa_dir=soa_dir,
)
sensu_output, sensu_status = sensu_message_status_for_jobs(
chronos_job_config=chronos_job_config,
service=service,
instance=instance,
cluster=cluster,
job_state_pairs=job_state_pairs)
monitoring_overrides = compose_monitoring_overrides_for_service(
chronos_job_config=chronos_job_config, soa_dir=soa_dir)
send_event(
service=service,
instance=instance,
monitoring_overrides=monitoring_overrides,
status_code=sensu_status,
message=sensu_output,
soa_dir=soa_dir,
)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
running_jobs = set(deployed_job_names(client))
expected_service_jobs = set([chronos_tools.compose_job_id(*job) for job in
chronos_tools.get_chronos_jobs_for_cluster(soa_dir=args.soa_dir)])
all_tmp_jobs = set(filter_tmp_jobs(filter_paasta_jobs(running_jobs)))
expired_tmp_jobs = set(filter_expired_tmp_jobs(client, all_tmp_jobs))
valid_tmp_jobs = all_tmp_jobs - expired_tmp_jobs
to_delete = running_jobs - expected_service_jobs - valid_tmp_jobs
task_responses = cleanup_tasks(client, to_delete)
task_successes = []
task_failures = []
for response in task_responses:
if isinstance(response[-1], Exception):
task_failures.append(response)
else:
task_successes.append(response)
job_responses = cleanup_jobs(client, to_delete)
job_successes = []
job_failures = []
for response in job_responses:
if isinstance(response[-1], Exception):
job_failures.append(response)
else:
job_successes.append(response)
try:
(service, instance) = chronos_tools.decompose_job_id(response[0])
send_event(
service=service,
instance=instance,
monitoring_overrides={},
soa_dir=soa_dir,
status_code=pysensu_yelp.Status.OK,
message="This instance was removed and is no longer supposed to be scheduled.",
)
except InvalidJobNameError:
# If we deleted some bogus job with a bogus jobid that could not be parsed,
# Just move on, no need to send any kind of paasta event.
pass
if len(to_delete) == 0:
print 'No Chronos Jobs to remove'
else:
if len(task_successes) > 0:
print format_list_output("Successfully Removed Tasks (if any were running) for:",
[job[0] for job in task_successes])
# if there are any failures, print and exit appropriately
if len(task_failures) > 0:
print format_list_output("Failed to Delete Tasks for:", [job[0] for job in task_failures])
if len(job_successes) > 0:
print format_list_output("Successfully Removed Jobs:", [job[0] for job in job_successes])
# if there are any failures, print and exit appropriately
if len(job_failures) > 0:
print format_list_output("Failed to Delete Jobs:", [job[0] for job in job_failures])
if len(job_failures) > 0 or len(task_failures) > 0:
sys.exit(1)
def main():
args = parse_args()
soa_dir = args.soa_dir
config = chronos_tools.load_chronos_config()
client = chronos_tools.get_chronos_client(config)
system_paasta_config = utils.load_system_paasta_config()
cluster = system_paasta_config.get_cluster()
running_jobs = set(deployed_job_names(client))
expected_service_jobs = {chronos_tools.compose_job_id(*job) for job in
chronos_tools.get_chronos_jobs_for_cluster(soa_dir=args.soa_dir)}
all_tmp_jobs = set(filter_tmp_jobs(filter_paasta_jobs(running_jobs)))
expired_tmp_jobs = set(filter_expired_tmp_jobs(client, all_tmp_jobs, cluster=cluster, soa_dir=soa_dir))
valid_tmp_jobs = all_tmp_jobs - expired_tmp_jobs
to_delete = running_jobs - expected_service_jobs - valid_tmp_jobs
task_responses = cleanup_tasks(client, to_delete)
task_successes = []
task_failures = []
for response in task_responses:
if isinstance(response[-1], Exception):
task_failures.append(response)
else:
task_successes.append(response)
job_responses = cleanup_jobs(client, to_delete)
job_successes = []
job_failures = []
for response in job_responses:
if isinstance(response[-1], Exception):
job_failures.append(response)
else:
job_successes.append(response)
try:
(service, instance) = chronos_tools.decompose_job_id(response[0])
monitoring_tools.send_event(
check_name=check_chronos_job_name(service, instance),
service=service,
overrides={},
soa_dir=soa_dir,
status=pysensu_yelp.Status.OK,
output="This instance was removed and is no longer supposed to be scheduled.",
)
except InvalidJobNameError:
# If we deleted some bogus job with a bogus jobid that could not be parsed,
# Just move on, no need to send any kind of paasta event.
pass
if len(to_delete) == 0:
paasta_print('No Chronos Jobs to remove')
else:
if len(task_successes) > 0:
paasta_print(format_list_output(
"Successfully Removed Tasks (if any were running) for:",
[job[0] for job in task_successes],
))
# if there are any failures, print and exit appropriately
if len(task_failures) > 0:
paasta_print(format_list_output("Failed to Delete Tasks for:", [job[0] for job in task_failures]))
if len(job_successes) > 0:
paasta_print(format_list_output("Successfully Removed Jobs:", [job[0] for job in job_successes]))
# if there are any failures, print and exit appropriately
if len(job_failures) > 0:
paasta_print(format_list_output("Failed to Delete Jobs:", [job[0] for job in job_failures]))
if len(job_failures) > 0 or len(task_failures) > 0:
sys.exit(1)
