def test_refresh_invalid_response(monkeypatch, tmpdir):
tokens.configure(dir=str(tmpdir), url='https://example.org')
tokens.manage('mytok', ['myscope'])
tokens.start() # this does not do anything..
response = MagicMock()
response.json.return_value = {'foo': 'bar'}
post = MagicMock()
post.return_value = response
monkeypatch.setattr('requests.post', post)
monkeypatch.setattr('tokens.read_credentials', lambda path: (VALID_USER_JSON, VALID_CLIENT_JSON))
with pytest.raises(tokens.InvalidTokenResponse) as exc_info:
tokens.get('mytok')
assert str(exc_info.value) == """Invalid token response: Expected a JSON object with keys "expires_in" and "access_token": 'expires_in'"""
# verify that we use a proper HTTP timeout..
post.assert_called_with('https://example.org',
data={'username': '******', 'scope': 'myscope', 'password': '******', 'grant_type': 'password'},
headers={'User-Agent': 'python-tokens/{}'.format(tokens.__version__)},
timeout=(1.25, 2.25), auth=('cid', 'sec'))
response.json.return_value = {'access_token': '', 'expires_in': 100}
with pytest.raises(tokens.InvalidTokenResponse) as exc_info:
tokens.get('mytok')
assert str(exc_info.value) == 'Invalid token response: Empty "access_token" value'
def get_token(url: str, scopes: str, credentials_dir: str) -> dict:
"""
Get access token info.
"""
tokens.configure(url=url, dir=credentials_dir)
tokens.manage('lizzy', [scopes])
tokens.start()
return tokens.get('lizzy')
def __init__(self, kafka_broker, metrics, callback_store, worker_thread_count=1, max_read_messages_per_cycle=1, use_oauth2_for_push=False):
threading.Thread.__init__(self, daemon=True)
self.use_oauth2_for_push = use_oauth2_for_push
self.callback_store = callback_store
self.worker_thread_count = worker_thread_count
self.max_read_messages_per_cycle = max_read_messages_per_cycle
self.kafka = kafka_broker
self.metrics = metrics
if use_oauth2_for_push:
tokens.manage('event-store', ['uid'])
tokens.start()
def __init__(self):
if os.getenv('CREDENTIALS_DIR') or os.getenv('OAUTH2_ACCESS_TOKENS'):
# no human user credentials if running as service
# CREDENTIALS_DIR means running on Taupage, OAUTH2_ACCESS_TOKENS means Docker locally
tokens.manage('costreport', ['uid'])
tokens.start()
self.on_taupage = True
elif os.getenv('TRAVIS'):
self.token = "i-am-on-travis-and-will-not-be-used"
self.on_taupage = False
else:
# human user
self.token = zign.api.get_token('costreport', ['uid'])
self.on_taupage = False
def use_berry_token(app_name):
tokens.configure()
tokens.manage(
app_name,
['nakadi.event_stream.read', 'nakadi.event_stream.write', 'nakadi.event_type.write', 'uid'])
tokens.start()
def _get_token():
try:
return tokens.get(app_name)
except Exception as e:
logging.error('Failed to get token for {}'.format(app_name), exc_info=e)
return ''
set_provider(_get_token)
def test_refresh_invalid_credentials(monkeypatch, tmpdir):
tokens.configure(dir=str(tmpdir), url='https://example.org')
tokens.manage('mytok', ['myscope'])
tokens.start() # this does not do anything..
with open(os.path.join(str(tmpdir), 'user.json'), 'w') as fd:
# missing password
json.dump({'application_username': '******'}, fd)
with open(os.path.join(str(tmpdir), 'client.json'), 'w') as fd:
json.dump({'client_id': 'cid', 'client_secret': 'sec'}, fd)
with pytest.raises(tokens.InvalidCredentialsError) as exc_info:
tokens.get('mytok')
assert str(exc_info.value) == "Invalid OAuth credentials: Missing key: 'application_password'"
def __init__(self,
kafka_broker,
metrics,
callback_store,
worker_thread_count=1,
max_read_messages_per_cycle=1,
use_oauth2_for_push=False):
threading.Thread.__init__(self, daemon=True)
self.use_oauth2_for_push = use_oauth2_for_push
self.callback_store = callback_store
self.worker_thread_count = worker_thread_count
self.max_read_messages_per_cycle = max_read_messages_per_cycle
self.kafka = kafka_broker
self.metrics = metrics
if use_oauth2_for_push:
tokens.manage('event-store', ['uid'])
tokens.start()
def configure(self, conf):
"""
Called after plugin is loaded to pass the [configuration] section in their plugin info file
:param conf: configuration dictionary
"""
# will use OAUTH2_ACCESS_TOKEN_URL environment variable by default
# will try to read application credentials from CREDENTIALS_DIR
tokens.configure()
token_configuration = conf.get('oauth2.tokens')
if token_configuration:
for part in token_configuration.split(':'):
token_name, scopes = tuple(part.split('=', 1))
tokens.manage(token_name, scopes.split(','))
tokens.manage('uid', ['uid'])
tokens.start()
def configure(self, conf):
"""
Called after plugin is loaded to pass the [configuration] section in their plugin info file
:param conf: configuration dictionary
"""
# will use OAUTH2_ACCESS_TOKEN_URL environment variable by default
# will try to read application credentials from CREDENTIALS_DIR
tokens.configure()
token_configuration = conf.get('oauth2.tokens')
if token_configuration:
for part in token_configuration.split(':'):
token_name, scopes = tuple(part.split('=', 1))
tokens.manage(token_name, scopes.split(','))
tokens.manage('uid', ['uid'])
tokens.start()
import requests
import tokens
from zmon_worker_monitor.zmon_worker.errors import ConfigurationError
from zmon_worker_monitor.builtins.plugins.distance_to_history import DistanceWrapper
from zmon_worker_monitor.adapters.ifunctionfactory_plugin import IFunctionFactoryPlugin, propartial
logger = logging.getLogger(__name__)
# will use OAUTH2_ACCESS_TOKEN_URL environment variable by default
# will try to read application credentials from CREDENTIALS_DIR
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
ONE_WEEK = 7 * 24 * 60 * 60
ONE_WEEK_AND_5MIN = ONE_WEEK + 5 * 60
DATAPOINTS_ENDPOINT = 'api/v1/datapoints/query'
class HistoryFactory(IFunctionFactoryPlugin):
def __init__(self):
super(HistoryFactory, self).__init__()
# fields from configuration
def configure(self, conf):
"""
Called after plugin is loaded to pass the [configuration] section in their plugin info file
def main():
argp = argparse.ArgumentParser(description='ZMON AWS Agent')
argp.add_argument('-e', '--entity-service', dest='entityservice')
argp.add_argument('-r', '--region', dest='region', default=None)
argp.add_argument('-j', '--json', dest='json', action='store_true')
argp.add_argument('--no-oauth2',
dest='disable_oauth2',
action='store_true',
default=False)
argp.add_argument('--postgresql-user',
dest='postgresql_user',
default=os.environ.get('AGENT_POSTGRESQL_USER'))
argp.add_argument('--postgresql-pass',
dest='postgresql_pass',
default=os.environ.get('AGENT_POSTGRESQL_PASS'))
args = argp.parse_args()
if not args.disable_oauth2:
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
logging.basicConfig(level=logging.INFO)
# 1. Determine region
if not args.region:
logger.info('Trying to figure out region..')
try:
response = requests.get(
'http://169.254.169.254/latest/meta-data/placement/availability-zone',
timeout=2)
except:
logger.exception(
'Region was not specified as a parameter and can not be fetched from instance meta-data!'
)
raise
region = response.text[:-1]
else:
region = args.region
logger.info('Using region: {}'.format(region))
logger.info('Entity service URL: %s', args.entityservice)
logger.info('Reading DNS data for hosted zones')
aws.populate_dns_data()
aws_account_id = aws.get_account_id(region)
infrastructure_account = 'aws:{}'.format(
aws_account_id) if aws_account_id else None
if not infrastructure_account:
logger.error(
'AWS agent: Cannot determine infrastructure account ID. Terminating!'
)
return
# 2. ZMON entities
token = None if args.disable_oauth2 else tokens.get('uid')
zmon_client = Zmon(args.entityservice,
token=token,
user_agent=get_user_agent())
query = {
'infrastructure_account': infrastructure_account,
'region': region,
'created_by': 'agent'
}
entities = zmon_client.get_entities(query)
# 3. Get running apps
apps = aws.get_running_apps(region, entities)
elbs = []
scaling_groups = []
rds = []
elasticaches = []
dynamodbs = []
sqs = []
new_entities = []
to_be_removed = []
if len(apps) > 0:
elbs = aws.get_running_elbs(region, infrastructure_account)
scaling_groups = aws.get_auto_scaling_groups(region,
infrastructure_account)
rds = aws.get_rds_instances(region, infrastructure_account, entities)
elasticaches = aws.get_elasticache_nodes(region,
infrastructure_account)
dynamodbs = aws.get_dynamodb_tables(region, infrastructure_account)
certificates = aws.get_certificates(region, infrastructure_account)
aws_limits = aws.get_limits(region, infrastructure_account, apps, elbs)
sqs = aws.get_sqs_queues(region, infrastructure_account, entities)
account_alias = aws.get_account_alias(region)
ia_entity = {
'type': 'local',
'infrastructure_account': infrastructure_account,
'account_alias': account_alias,
'region': region,
'id': 'aws-ac[{}:{}]'.format(infrastructure_account, region),
'created_by': 'agent',
}
application_entities = aws.get_apps_from_entities(apps,
infrastructure_account,
region)
if args.postgresql_user and args.postgresql_pass:
postgresql_clusters = zmon_client.get_entities({
'infrastructure_account':
infrastructure_account,
'region':
region,
'type':
'postgresql_cluster'
})
postgresql_databases = postgresql.get_databases_from_clusters(
postgresql_clusters, infrastructure_account, region,
args.postgresql_user, args.postgresql_pass)
else:
# Pretend the list of DBs is empty, but also make sure we don't remove
# any pre-existing database entities because we don't know about them.
postgresql_databases = []
entities = [
e for e in entities if e.get('type') != 'postgresql_database'
]
current_entities = (elbs + scaling_groups + apps + application_entities +
rds + postgresql_databases + elasticaches + dynamodbs +
certificates + sqs)
current_entities.append(aws_limits)
current_entities.append(ia_entity)
# 4. Removing misssing entities
existing_ids = get_existing_ids(entities)
current_entities_ids = {e['id'] for e in current_entities}
to_be_removed, delete_error_count = remove_missing_entities(
existing_ids, current_entities_ids, zmon_client, json=args.json)
logger.info(
'Found {} removed entities from {} entities ({} failed)'.format(
len(new_entities), len(current_entities), delete_error_count))
# 5. Get new/updated entities
new_entities, add_error_count = add_new_entities(current_entities,
entities,
zmon_client,
json=args.json)
logger.info('Found {} new entities from {} entities ({} failed)'.format(
len(new_entities), len(current_entities), add_error_count))
# 6. Always add Local entity
if not args.json:
ia_entity['errors'] = {
'delete_count': delete_error_count,
'add_count': add_error_count
}
try:
zmon_client.add_entity(ia_entity)
except:
logger.exception(
'Failed to add Local entity: {}'.format(ia_entity))
types = {
e['type']: len([t for t in new_entities if t['type'] == e['type']])
for e in new_entities
}
for t, v in types.items():
logger.info('Found {} new entities of type: {}'.format(v, t))
# Check if it is a dry run!
if args.json:
d = {
'applications': application_entities,
'apps': apps,
'dynamodb': dynamodbs,
'elbs': elbs,
'elc': elasticaches,
'rds': rds,
'certificates': certificates,
'aws_limits': aws_limits,
'sqs_queues': sqs,
'new_entities': new_entities,
'to_be_removed': to_be_removed,
}
print(json.dumps(d, indent=4))
import json
from prometheus_client.parser import text_string_to_metric_families
from collections import defaultdict
from zmon_worker_monitor.zmon_worker.errors import HttpError
from requests.adapters import HTTPAdapter
from zmon_worker_monitor.adapters.ifunctionfactory_plugin import IFunctionFactoryPlugin, propartial
import tokens
# will use OAUTH2_ACCESS_TOKEN_URL environment variable by default
# will try to read application credentials from CREDENTIALS_DIR
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
logger = logging.getLogger('zmon-worker.http-function')
class HttpFactory(IFunctionFactoryPlugin):
def __init__(self):
super(HttpFactory, self).__init__()
def configure(self, conf):
"""
Called after plugin is loaded to pass the [configuration] section in their plugin info file
:param conf: configuration dictionary
"""
return
def main():
argp = argparse.ArgumentParser(description='ZMON AWS Agent')
argp.add_argument('-e', '--entity-service', dest='entityservice')
argp.add_argument('-r', '--region', dest='region', default=None)
argp.add_argument('-j', '--json', dest='json', action='store_true')
argp.add_argument('-t',
'--tracer',
dest='tracer',
default=os.environ.get('OPENTRACING_TRACER', 'noop'))
argp.add_argument('--no-oauth2',
dest='disable_oauth2',
action='store_true',
default=False)
argp.add_argument('--postgresql-user',
dest='postgresql_user',
default=os.environ.get('AGENT_POSTGRESQL_USER'))
argp.add_argument('--postgresql-pass',
dest='postgresql_pass',
default=os.environ.get('AGENT_POSTGRESQL_PASS'))
args = argp.parse_args()
if not args.disable_oauth2:
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
init_opentracing_tracer(args.tracer)
root_span = opentracing.tracer.start_span(
operation_name='aws_entity_discovery')
with root_span:
logging.basicConfig(level=logging.INFO)
# 0. Fetch extra data for entities
entity_extras = {}
for ex in os.getenv('EXTRA_ENTITY_FIELDS', '').split(','):
if '=' not in ex:
continue
k, v = ex.split('=', 1)
if k and v:
entity_extras[k] = v
# 1. Determine region
if not args.region:
logger.info('Trying to figure out region..')
try:
response = requests.get(
'http://169.254.169.254/latest/meta-data/placement/availability-zone',
timeout=2)
except Exception:
root_span.set_tag('error', True)
root_span.log_kv({'exception': traceback.format_exc()})
logger.exception(
'Region was not specified as a parameter and' +
'can not be fetched from instance meta-data!')
raise
region = response.text[:-1]
else:
region = args.region
root_span.set_tag('region', region)
logger.info('Using region: {}'.format(region))
logger.info('Entity service URL: %s', args.entityservice)
logger.info('Reading DNS data for hosted zones')
aws.populate_dns_data()
aws_account_id = aws.get_account_id(region)
infrastructure_account = 'aws:{}'.format(
aws_account_id) if aws_account_id else None
if not infrastructure_account:
logger.error(
'AWS agent: Cannot determine infrastructure account ID. Terminating!'
)
return
root_span.set_tag('account', infrastructure_account)
# 2. ZMON entities
if not args.disable_oauth2:
token = os.getenv('ZMON_TOKEN', None) or tokens.get('uid')
zmon_client = Zmon(args.entityservice,
token=token,
user_agent=get_user_agent())
query = {
'infrastructure_account': infrastructure_account,
'region': region,
'created_by': 'agent'
}
entities = zmon_client.get_entities(query)
# 3. Get running apps
apps = aws.get_running_apps(region, entities)
elbs = []
scaling_groups = []
rds = []
elasticaches = []
dynamodbs = []
sqs = []
new_entities = []
to_be_removed = []
if len(apps) > 0:
elbs = aws.get_running_elbs(region, infrastructure_account)
scaling_groups = aws.get_auto_scaling_groups(
region, infrastructure_account)
rds = aws.get_rds_instances(region, infrastructure_account,
entities)
elasticaches = aws.get_elasticache_nodes(region,
infrastructure_account)
dynamodbs = aws.get_dynamodb_tables(region, infrastructure_account)
certificates = aws.get_certificates(region, infrastructure_account)
aws_limits = aws.get_limits(region, infrastructure_account, apps,
elbs, entities)
sqs = aws.get_sqs_queues(region, infrastructure_account, entities)
postgresql_clusters = postgresql.get_postgresql_clusters(
region, infrastructure_account, scaling_groups, apps)
account_alias = aws.get_account_alias(region)
ia_entity = {
'type': 'local',
'infrastructure_account': infrastructure_account,
'account_alias': account_alias,
'region': region,
'id': 'aws-ac[{}:{}]'.format(infrastructure_account, region),
'created_by': 'agent',
}
account_alias_prefix = os.getenv('ACCOUNT_ALIAS_PREFIX', None)
owner = account_alias
if account_alias_prefix:
owner = owner.replace(account_alias_prefix, '', 1)
root_span.set_tag('team', owner)
application_entities = aws.get_apps_from_entities(
apps, infrastructure_account, region)
if args.postgresql_user and args.postgresql_pass:
postgresql_databases = postgresql.get_databases_from_clusters(
postgresql_clusters, infrastructure_account, region,
args.postgresql_user, args.postgresql_pass)
else:
# Pretend the list of DBs is empty, but also make sure we don't remove
# any pre-existing database entities because we don't know about them.
postgresql_databases = []
entities = [
e for e in entities if e.get('type') != 'postgresql_database'
]
current_entities = (elbs + scaling_groups + apps +
application_entities + rds + postgresql_databases +
postgresql_clusters + elasticaches + dynamodbs +
certificates + sqs)
current_entities.append(aws_limits)
current_entities.append(ia_entity)
for entity in current_entities:
entity.update(entity_extras)
# 4. Removing misssing entities
existing_ids = get_existing_ids(entities)
current_entities_ids = {e['id'] for e in current_entities}
to_be_removed, delete_error_count = remove_missing_entities(
existing_ids, current_entities_ids, zmon_client, json=args.json)
root_span.log_kv({'total_entitites': str(len(current_entities))})
root_span.log_kv({'removed_entities': str(len(to_be_removed))})
logger.info(
'Found {} removed entities from {} entities ({} failed)'.format(
len(to_be_removed), len(current_entities), delete_error_count))
# 5. Get new/updated entities
new_entities, add_error_count = add_new_entities(current_entities,
entities,
zmon_client,
json=args.json)
root_span.log_kv({'new_entities': str(len(new_entities))})
logger.info(
'Found {} new entities from {} entities ({} failed)'.format(
len(new_entities), len(current_entities), add_error_count))
# 6. Always add Local entity
if not args.json:
ia_entity['errors'] = {
'delete_count': delete_error_count,
'add_count': add_error_count
}
update_local_entity(zmon_client, ia_entity)
types = {
e['type']: len([t for t in new_entities if t['type'] == e['type']])
for e in new_entities
}
for t, v in types.items():
logger.info('Found {} new entities of type: {}'.format(v, t))
# Check if it is a dry run!
if args.json:
d = {
'applications': application_entities,
'apps': apps,
'dynamodb': dynamodbs,
'elbs': elbs,
'elc': elasticaches,
'rds': rds,
'certificates': certificates,
'aws_limits': aws_limits,
'sqs_queues': sqs,
'new_entities': new_entities,
'to_be_removed': to_be_removed,
'posgresql_clusters': postgresql_clusters
}
print(json.dumps(d, indent=4))
def main():
argp = argparse.ArgumentParser(description='ZMON AWS Agent')
argp.add_argument('-e', '--entity-service', dest='entityservice')
argp.add_argument('-r', '--region', dest='region', default=None)
argp.add_argument('-j', '--json', dest='json', action='store_true')
argp.add_argument('-t', '--tracer', dest='tracer', default=os.environ.get('OPENTRACING_TRACER', 'noop'))
argp.add_argument('--no-oauth2', dest='disable_oauth2', action='store_true', default=False)
argp.add_argument('--postgresql-user', dest='postgresql_user', default=os.environ.get('AGENT_POSTGRESQL_USER'))
argp.add_argument('--postgresql-pass', dest='postgresql_pass', default=os.environ.get('AGENT_POSTGRESQL_PASS'))
args = argp.parse_args()
if not args.disable_oauth2:
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
init_opentracing_tracer(args.tracer)
root_span = opentracing.tracer.start_span(operation_name='aws_entity_discovery')
with root_span:
logging.basicConfig(level=logging.INFO)
# 0. Fetch extra data for entities
entity_extras = {}
for ex in os.getenv('EXTRA_ENTITY_FIELDS', '').split(','):
if '=' not in ex:
continue
k, v = ex.split('=', 1)
if k and v:
entity_extras[k] = v
# 1. Determine region
if not args.region:
logger.info('Trying to figure out region..')
try:
response = requests.get('http://169.254.169.254/latest/meta-data/placement/availability-zone',
timeout=2)
except Exception:
root_span.set_tag('error', True)
root_span.log_kv({'exception': traceback.format_exc()})
logger.exception('Region was not specified as a parameter and' +
'can not be fetched from instance meta-data!')
raise
region = response.text[:-1]
else:
region = args.region
root_span.set_tag('region', region)
logger.info('Using region: {}'.format(region))
logger.info('Entity service URL: %s', args.entityservice)
logger.info('Reading DNS data for hosted zones')
aws.populate_dns_data()
aws_account_id = aws.get_account_id(region)
infrastructure_account = 'aws:{}'.format(aws_account_id) if aws_account_id else None
if not infrastructure_account:
logger.error('AWS agent: Cannot determine infrastructure account ID. Terminating!')
return
root_span.set_tag('account', infrastructure_account)
# 2. ZMON entities
if not args.disable_oauth2:
token = os.getenv('ZMON_TOKEN', None) or tokens.get('uid')
zmon_client = Zmon(args.entityservice, token=token, user_agent=get_user_agent())
query = {'infrastructure_account': infrastructure_account, 'region': region, 'created_by': 'agent'}
entities = zmon_client.get_entities(query)
# 3. Get running apps
apps = aws.get_running_apps(region, entities)
elbs = []
scaling_groups = []
elastigroups = []
certificates = []
rds = []
elasticaches = []
dynamodbs = []
sqs = []
postgresql_clusters = []
aws_limits = []
new_entities = []
to_be_removed = []
if len(apps) > 0:
elbs = aws.get_running_elbs(region, infrastructure_account)
scaling_groups = aws.get_auto_scaling_groups(region, infrastructure_account)
elastigroups = elastigroup.get_elastigroup_entities(region, infrastructure_account)
rds = aws.get_rds_instances(region, infrastructure_account, entities)
elasticaches = aws.get_elasticache_nodes(region, infrastructure_account)
dynamodbs = aws.get_dynamodb_tables(region, infrastructure_account)
certificates = aws.get_certificates(region, infrastructure_account)
aws_limits = aws.get_limits(region, infrastructure_account, apps, elbs, entities)
sqs = aws.get_sqs_queues(region, infrastructure_account, entities)
postgresql_clusters = postgresql.get_postgresql_clusters(region, infrastructure_account,
scaling_groups, apps)
account_alias = aws.get_account_alias(region)
ia_entity = {
'type': 'local',
'infrastructure_account': infrastructure_account,
'account_alias': account_alias,
'region': region,
'id': 'aws-ac[{}:{}]'.format(infrastructure_account, region),
'created_by': 'agent',
}
account_alias_prefix = os.getenv('ACCOUNT_ALIAS_PREFIX', None)
owner = account_alias
if account_alias_prefix:
owner = owner.replace(account_alias_prefix, '', 1)
root_span.set_tag('team', owner)
application_entities = aws.get_apps_from_entities(apps, infrastructure_account, region)
if args.postgresql_user and args.postgresql_pass:
postgresql_databases = postgresql.get_databases_from_clusters(postgresql_clusters,
infrastructure_account,
region,
args.postgresql_user,
args.postgresql_pass)
else:
# Pretend the list of DBs is empty, but also make sure we don't remove
# any pre-existing database entities because we don't know about them.
postgresql_databases = []
entities = [e for e in entities if e.get('type') != 'postgresql_database']
current_entities = (
elbs + scaling_groups + elastigroups + apps + application_entities +
rds + postgresql_databases + postgresql_clusters + elasticaches + dynamodbs +
certificates + sqs)
current_entities.append(aws_limits)
current_entities.append(ia_entity)
for entity in current_entities:
entity.update(entity_extras)
# 4. Removing misssing entities
existing_ids = get_existing_ids(entities)
current_entities_ids = {e['id'] for e in current_entities}
to_be_removed, delete_error_count = remove_missing_entities(
existing_ids, current_entities_ids, zmon_client, json=args.json)
root_span.log_kv({'total_entitites': str(len(current_entities))})
root_span.log_kv({'removed_entities': str(len(to_be_removed))})
logger.info('Found {} removed entities from {} entities ({} failed)'.format(
len(to_be_removed), len(current_entities), delete_error_count))
# 5. Get new/updated entities
new_entities, add_error_count = add_new_entities(current_entities, entities, zmon_client, json=args.json)
root_span.log_kv({'new_entities': str(len(new_entities))})
logger.info('Found {} new entities from {} entities ({} failed)'.format(
len(new_entities), len(current_entities), add_error_count))
# 6. Always add Local entity
if not args.json:
ia_entity['errors'] = {'delete_count': delete_error_count, 'add_count': add_error_count}
update_local_entity(zmon_client, ia_entity)
types = {e['type']: len([t for t in new_entities if t['type'] == e['type']]) for e in new_entities}
for t, v in types.items():
logger.info('Found {} new entities of type: {}'.format(v, t))
# Check if it is a dry run!
if args.json:
d = {
'applications': application_entities,
'apps': apps,
'elastigroups': elastigroups,
'dynamodb': dynamodbs,
'elbs': elbs,
'elc': elasticaches,
'rds': rds,
'certificates': certificates,
'aws_limits': aws_limits,
'sqs_queues': sqs,
'new_entities': new_entities,
'to_be_removed': to_be_removed,
'posgresql_clusters': postgresql_clusters
}
print(json.dumps(d, indent=4))
def main():
argp = argparse.ArgumentParser(description='ZMon AWS Agent')
argp.add_argument('-e', '--entity-service', dest='entityservice')
argp.add_argument('-r', '--region', dest='region', default=None)
argp.add_argument('-j', '--json', dest='json', action='store_true')
argp.add_argument('--no-oauth2', dest='disable_oauth2', action='store_true', default=False)
args = argp.parse_args()
if not args.disable_oauth2:
tokens.configure()
tokens.manage('uid', ['uid'])
tokens.start()
logging.basicConfig(level=logging.INFO)
if not args.region:
logging.info("Trying to figure out region...")
try:
response = requests.get('http://169.254.169.254/latest/meta-data/placement/availability-zone', timeout=2)
except:
logging.error("Region was not specified as a parameter and can not be fetched from instance meta-data!")
raise
region = response.text[:-1]
else:
region = args.region
logging.info("Using region: {}".format(region))
logging.info("Entity service url: %s", args.entityservice)
apps = get_running_apps(region)
if len(apps) > 0:
infrastructure_account = apps[0]['infrastructure_account']
elbs = get_running_elbs(region, infrastructure_account)
scaling_groups = get_auto_scaling_groups(region, infrastructure_account)
rds = get_rds_instances(region, infrastructure_account)
elasticaches = get_elasticache_nodes(region, infrastructure_account)
dynamodbs = get_dynamodb_tables(region, infrastructure_account)
else:
elbs = []
scaling_groups = []
rds = []
if args.json:
d = {'apps': apps, 'elbs': elbs, 'rds': rds, 'elc': elasticaches, 'dynamodb': dynamodbs}
print(json.dumps(d))
else:
if infrastructure_account is not None:
account_alias = get_account_alias(region)
ia_entity = {"type": "local",
"infrastructure_account": infrastructure_account,
"account_alias": account_alias,
"region": region,
"id": "aws-ac[{}:{}]".format(infrastructure_account, region),
"created_by": "agent"}
application_entities = get_apps_from_entities(apps, infrastructure_account, region)
current_entities = []
for e in elbs:
current_entities.append(e["id"])
for e in scaling_groups:
current_entities.append(e["id"])
for a in apps:
current_entities.append(a["id"])
for a in application_entities:
current_entities.append(a["id"])
for a in rds:
current_entities.append(a["id"])
for a in elasticaches:
current_entities.append(a["id"])
for a in dynamodbs:
current_entities.append(a["id"])
current_entities.append(ia_entity["id"])
headers = {'Content-Type': 'application/json'}
if not args.disable_oauth2:
token = os.getenv('ZMON_AGENT_TOKEN', tokens.get('uid'))
logging.info("Adding oauth2 token to requests {}...{}".format(token[:1], token[-1:]))
headers.update({'Authorization': 'Bearer {}'.format(token)})
# removing all entities
query = {'infrastructure_account': infrastructure_account, 'region': region, 'created_by': 'agent'}
r = requests.get(args.entityservice,
params={'query': json.dumps(query)}, headers=headers, timeout=10)
entities = r.json()
existing_entities = {}
to_remove = []
for e in entities:
existing_entities[e['id']] = e
if not e["id"] in current_entities:
to_remove.append(e["id"])
if os.getenv('zmon_user'):
auth = (os.getenv('zmon_user'), os.getenv('zmon_password', ''))
else:
auth = None
for e in to_remove:
logging.info("removing instance: {}".format(e))
r = requests.delete(args.entityservice + "{}/".format(e), auth=auth, headers=headers, timeout=3)
logging.info("...%s", r.status_code)
def put_entity(entity_type, entity):
logging.info("Adding {} entity: {}".format(entity_type, entity['id']))
r = requests.put(args.entityservice, auth=auth,
data=json.dumps(entity, default=json_serial),
headers=headers, timeout=3)
logging.info("...%s", r.status_code)
put_entity('LOCAL', ia_entity)
for instance in apps:
put_entity('instance', instance)
for asg in scaling_groups:
put_entity('Auto Scaling group', asg)
for elb in elbs:
put_entity('elastic load balancer', elb)
for db in rds:
put_entity('RDS instance', db)
# merge here or we loose it on next pull
for app in application_entities:
if app['id'] in existing_entities:
ex = existing_entities[app['id']]
if 'scalyr_ts_id' in ex:
app['scalyr_ts_id'] = ex['scalyr_ts_id']
for app in application_entities:
put_entity('application', app)
for elasticache in elasticaches:
put_entity('elasticache', elasticache)
for dynamodb in dynamodbs:
put_entity('dynamodb', dynamodb)
