def _register(self):
if self.name in all_gauges.keys():
logger.warning("%s already registered, reregistering" % (self.name,))
REGISTRY.unregister(all_gauges.pop(self.name))
REGISTRY.register(self)
all_gauges[self.name] = self
def start_exporter(config, port, interval):
""" run the exporter every <interval> seconds """
REGISTRY.register(NovaCollector(config))
start_http_server(port)
while True:
generate_latest(REGISTRY)
time.sleep(30)
def main():
"""
Main method
"""
args = create_parser().parse_args()
log_format = '%(asctime)s %(message)s'
logging_args = dict(format=log_format,
level=args.log_level)
if args.log_file:
logging_args['filename'] = args.log_file
logging_args['filemode'] = 'a'
logging.basicConfig(**logging_args)
scheme = "https" if args.use_tls else "http"
use_ts = args.use_device_data_timestamp
collector = SunPowerPVSupervisorCollector(hostname=args.hostname,
port=args.port,
scheme=scheme,
timeout=args.timeout,
use_device_data_timestamp=use_ts,
)
logging.info("Listening on port %d...", args.listen_on)
start_http_server(args.listen_on)
REGISTRY.register(collector)
# Sleep indefinitely until we receive a signal
while True:
time.sleep(10)
def main():
parser = create_parser()
args = parser.parse_args()
# log configuration
loginipath = os.path.join(os.path.dirname(__file__)) + '/' + "logconf.ini"
logging.config.fileConfig(
loginipath,
defaults={'logfilename': args.logfile}
)
logger = logging.getLogger("zvmExporter")
# split address and port
addr_rx = re.compile(
r'(?P<addr>[a-zA-Z0-9][a-zA-Z0-9\-]*(\.[a-zA-Z0-9][a-zA-Z0-9\-]*)+)'
r'(:(?P<port>\d+))?')
match = addr_rx.match(args.server)
if match:
xcat_addr = match.group('addr')
xcat_port = match.group('port') or '443'
else:
logger.info("Invalid address")
return 1
logger.info("Program started")
# start collector
REGISTRY.register(ZVMCollector(args.zhcpnode, args.username,
args.password, xcat_addr, xcat_port,
args.cert))
start_http_server(args.port)
while True:
sleep(1)
def exporter_start():
print('starting server http://{}:{}/metrics'.format(
EXPORTER_LISTEN_HOST, EXPORTER_LISTEN_PORT))
REGISTRY.register(CustomCollector())
start_http_server(EXPORTER_LISTEN_PORT, addr=EXPORTER_LISTEN_HOST)
while True:
time.sleep(5)
def main():
try:
args = parse_args()
port = int(args.port)
REGISTRY.register(AzureStatusCollector())
start_http_server(port)
while True:
time.sleep(1)
except KeyboardInterrupt:
print(" Interrupted")
exit(0)
def main():
try:
args = parse_args()
port = int(args.port)
REGISTRY.register(CouchbaseCollector(args.couchbase))
start_http_server(port)
print "Serving at port: ", port
while True: time.sleep(1)
except KeyboardInterrupt:
print(" Interrupted")
exit(0)
def main():
try:
args = parse_args()
port = int(args.port)
REGISTRY.register(JenkinsCollector(args.jenkins, args.user, args.password, args.insecure))
start_http_server(port)
print("Polling {}. Serving at port: {}".format(args.jenkins, port))
while True:
time.sleep(1)
except KeyboardInterrupt:
print(" Interrupted")
exit(0)
def register_cache(cache_type, cache_name, cache):
# Check if the metric is already registered. Unregister it, if so.
# This usually happens during tests, as at runtime these caches are
# effectively singletons.
metric_name = "cache_%s_%s" % (cache_type, cache_name)
if metric_name in collectors_by_name.keys():
REGISTRY.unregister(collectors_by_name[metric_name])
class CacheMetric(object):
hits = 0
misses = 0
evicted_size = 0
def inc_hits(self):
self.hits += 1
def inc_misses(self):
self.misses += 1
def inc_evictions(self, size=1):
self.evicted_size += size
def describe(self):
return []
def collect(self):
try:
if cache_type == "response_cache":
response_cache_size.labels(cache_name).set(len(cache))
response_cache_hits.labels(cache_name).set(self.hits)
response_cache_evicted.labels(cache_name).set(self.evicted_size)
response_cache_total.labels(cache_name).set(self.hits + self.misses)
else:
cache_size.labels(cache_name).set(len(cache))
cache_hits.labels(cache_name).set(self.hits)
cache_evicted.labels(cache_name).set(self.evicted_size)
cache_total.labels(cache_name).set(self.hits + self.misses)
except Exception as e:
logger.warn("Error calculating metrics for %s: %s", cache_name, e)
raise
yield GaugeMetricFamily("__unused", "")
metric = CacheMetric()
REGISTRY.register(metric)
caches_by_name[cache_name] = cache
collectors_by_name[metric_name] = metric
return metric
def install_colletor(self):
class Collector(object):
def collect(self):
try:
ret = []
for c in kvmagent.metric_collectors:
ret.extend(c())
return ret
except Exception as e:
content = traceback.format_exc()
err = '%s\n%s\n' % (str(e), content)
logger.warn(err)
return []
REGISTRY.register(Collector())
def main():
try:
args = parse_args()
port = int(args.port)
while True:
try:
if requests.get(args.odl_inventory):
REGISTRY.register(OpenDaylightCollector(args.opendaylight, args.odl_inventory))
start_http_server(port)
print "Polling data from OpenDaylight: %s. Starting OpenDaylight exporter on port: %s" % (args.opendaylight, port)
while True:
time.sleep(1)
except ConnectionError:
print "OpenDaylight is either not running or it is unreachable."
except KeyboardInterrupt:
print(" Interrupted")
exit(0)
def run_server(test, port):
"""
This script provides monitoring information about the postgraas server.
"""
logging.basicConfig(level=logging.INFO, format="%(asctime)s %(name)-12s %(levelname)-8s %(message)s")
config = cfg.get_config()
collector = CustomCollector(config)
if test:
click.echo("TEST MODE")
for metric in collector.collect():
for sample in metric.samples:
click.echo(sample)
else:
click.echo("Running web server at port {}".format(port))
REGISTRY.register(collector)
app = Flask(__name__)
app.register_blueprint(blueprint)
app.run(host='0.0.0.0', port=port, threaded=True)
type=int,
default=9683,
help='Collector http port, default is 9683')
parser.add_argument(
'--can-reset-counter',
dest='can_reset_counter',
help='Will reset counter as required when maxed out. Can also be \
set with env variable CAN_RESET_COUNTER',
action='store_true')
parser.add_argument(
'--from-file',
action='store',
dest='input_file',
help='Read a file containing the output of ibqueryerrors, if left \
empty, ibqueryerrors will be launched as needed by this collector')
parser.add_argument(
'--node-name-map',
action='store',
dest='node_name_map',
help='Node name map used by ibqueryerrors. Ccan also be set with env \
var NODE_NAME_MAP')
args = parser.parse_args()
start_http_server(args.port)
REGISTRY.register(
InfinibandCollector(args.can_reset_counter, args.input_file,
args.node_name_map))
while True:
time.sleep(1)
data["ports"][number]["packetsRxErrors"])
self.metric['packetsTxErrors'].add_metric(
[data["device"],
str(data["ports"][number]['port'])],
data["ports"][number]["packetsTxErrors"])
tables = device.get_table_statistic()["statistics"][0]
tables_length = len(tables['table'])
for number in range(0, tables_length):
if tables["table"][number]["activeEntries"] > 0:
self.metric['packetsMatched'].add_metric([
tables["device"],
str(tables["table"][number]['tableId'])
], tables["table"][number]["packetsMatched"])
self.metric['packetsLookedUp'].add_metric([
tables["device"],
str(tables["table"][number]['tableId'])
], tables["table"][number]["packetsLookedUp"])
for metric in self.metric.values():
yield metric
if __name__ == "__main__":
REGISTRY.register(prometheusCollector())
start_http_server(9091)
while True:
time.sleep(1)
def gyroscope_metric():
raw = sense.get_gyroscope_raw()
family = GaugeMetricFamily(name='sense_hat_gyroscope_raw',
documentation='radians/second',
labels=["axis"])
family.add_metric(["x"], raw["x"])
family.add_metric(["y"], raw["y"])
family.add_metric(["z"], raw["z"])
return family
def accelerometer_metric():
raw = sense.get_accelerometer_raw()
family = GaugeMetricFamily(
name='sense_hat_accelerometer_raw',
documentation='Measured by Raspberry Pi Sense Hat',
labels=["axis"])
family.add_metric(["x"], raw["x"])
family.add_metric(["y"], raw["y"])
family.add_metric(["z"], raw["z"])
return family
REGISTRY.register(SenseHatCollector())
# Registered at https://github.com/prometheus/prometheus/wiki/Default-port-allocations
start_http_server(9607)
while True:
time.sleep(60)
# kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group clk --describe | grep dtss-platform-cl | awk 'BEGIN{sum=0}{sum+=$5}END{print sum}'
lags['dtss_platform_clk'] = cmdout("kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group {} "
"--describe | grep {} | awk 'BEGIN{{sum=0}}{{sum+=$5}}END{{print "
"sum}}'".format('clk', 'dtss-platform-clk'))
lags['dtss_realtime_base'] = cmdout("kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group "
"{} --describe | grep {} | awk 'BEGIN{{sum=0}}{{"
"sum+=$5}}END{{print sum}}'".format('base', 'dtss-realtime-base'))
lags['dtss_stats_cvt'] = cmdout("kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group "
"{} --describe | grep {} | awk 'BEGIN{{sum=0}}{{"
"sum+=$5}}END{{print sum}}'".format('cvt', 'dtss-stats-cvt'))
lags['dtss_stats_cvt_my'] = cmdout("kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group "
"{} --describe | grep {} | awk 'BEGIN{{sum=0}}{{"
"sum+=$5}}END{{print sum}}'".format('my', 'dtss-stats-cvt'))
lags['dtss_realtime_account'] = cmdout("kafka-consumer-groups.sh --bootstrap-server 172.28.9.1:9092 --group "
"{} --describe | grep {} | awk 'BEGIN{{sum=0}}{{"
"sum+=$5}}END{{print sum}}'".format('account', 'dtss-realtime-account'))
for key in lags:
# __init__(self, name, documentation, value=None, labels=None, unit=''):
lags_gauges[key] = GaugeMetricFamily('td_{}'.format(key), 'metric of {}'.format(key), value=None, labels=['proj', 'service'])
lags_gauges[key].add_metric(['td', 'kafka'], lags[key])
for metric in lags_gauges:
yield lags_gauges[metric]
if __name__ == '__main__':
start_http_server(23333)
REGISTRY.register(lagCollector())
while True:
time.sleep(10)
if 'lastSuccessfulBuildId' in self.data[repo]:
lsbi = CounterMetricFamily(
'dronedb_repo_last_successful_build_id',
'Last successful build id',
labels=["repo"])
lsbi.add_metric([repo],
str(self.data[repo]['lastSuccessfulBuildId']))
yield (lsbi)
if 'lastBuildId' in self.data[repo]:
lbi = CounterMetricFamily('dronedb_repo_last_build_id',
'Last build id',
labels=["repo"])
lbi.add_metric([repo], str(self.data[repo]['lastBuildId']))
yield (lbi)
self.drone.close_db()
return (self.data)
if __name__ == "__main__":
args = setup_parser()
LOG.info('Starting dronedb exporter on http://localhost:%s' %
args.listen_port)
REGISTRY.register(
DroneCollector(args.host, args.dbport, args.database, args.username,
args.password))
start_http_server(int(args.listen_port))
while True:
time.sleep(1)
Imports
:time: control sleep time in main function
"""
import time
from datetime import datetime, timedelta
from prometheus_client import PLATFORM_COLLECTOR
from prometheus_client import PROCESS_COLLECTOR
from prometheus_client import start_http_server
from prometheus_client.core import CounterMetricFamily
from prometheus_client.core import REGISTRY
import lib.wg_parser as wg_parser
# unregister not used metrics
# pylint: disable=protected-access
REGISTRY.unregister(PROCESS_COLLECTOR)
REGISTRY.unregister(PLATFORM_COLLECTOR)
REGISTRY.unregister(
REGISTRY._names_to_collectors["python_gc_objects_collected_total"])
# pylint: disable=too-few-public-methods
class CollectSendBytesTotal:
"""
Custom collector class for bytes sent to wireguard peer
"""
def __init__(self, wgparser: wg_parser.WGParser):
self.parser = wgparser
def collect(self):
"""
except KeyError:
logging.error("Could not retrieve metrics from: " + self.metrics)
logging.error("Check argument sonar_metrics")
if __name__ == "__main__":
parser = configargparse.ArgumentParser()
parser.add_argument('--sonar_url', type=str, required=True, env_var='sonar_url')
parser.add_argument('--sonar_metrics', type=str, env_var='sonar_metrics', default='ncloc,coverage')
parser.add_argument('--sonar_user', type=str, required=True, env_var='sonar_user')
parser.add_argument('--sonar_password', type=str, required=True, env_var='sonar_password')
parser.add_argument('--run_once', action='store_true')
args = parser.parse_args()
logging.basicConfig(level=logging.DEBUG, format='%(asctime)s %(message)s')
requests.packages.urllib3.disable_warnings(InsecureRequestWarning)
REGISTRY.register(SonarCollector(args.sonar_url, args.sonar_user, args.sonar_password, args.sonar_metrics))
if args.run_once:
for x in REGISTRY.collect():
logging.info(x)
for y in x.samples:
logging.info(y)
sys.exit("runonce")
start_http_server(9118)
while True:
time.sleep(1)
logger = logging.getLogger(__name__)
try:
config = Config()
except ValueError as e:
logger.critical(e)
logger.critical("Invalid configuration. Exiting.")
sys.exit(1)
logger.info("Starting Kibana Prometheus exporter version %s\n" %
config.version + config.description())
REGISTRY.register(
KibanaCollector(
config.kibana_url,
kibana_login=config.kibana_login,
kibana_password=config.kibana_password,
ignore_ssl=config.ignore_ssl,
))
try:
start_http_server(config.listen_port)
except PermissionError as e:
logger.critical("Cannot bind to port %s. Permission denied.",
config.listen_port)
sys.exit(2)
loop = asyncio.new_event_loop()
try:
loop.run_forever()
except KeyboardInterrupt:
def main():
signal.signal(signal.SIGTERM, signal_handler)
parser = argparse.ArgumentParser(
description='Export ES query results to Prometheus.')
parser.add_argument(
'-e',
'--es-cluster',
default='localhost',
help='addresses of nodes in a Elasticsearch cluster to run queries on. '
'Nodes should be separated by commas e.g. es1,es2. '
'Ports can be provided if non-standard (9200) e.g. es1:9999. '
'Include the scheme for non-http nodes e.g. https://es1:9200. '
'--ca-certs must be provided for SSL certificate verification. '
'(default: localhost)')
parser.add_argument(
'--cluster-name',
help='cluster name, if given a value, it will be added to labels.')
parser.add_argument(
'--ca-certs',
help='path to a CA certificate bundle. '
'Can be absolute, or relative to the current working directory. '
'If not specified, SSL certificate verification is disabled.')
parser.add_argument(
'--client-cert',
help='path to a SSL client certificate. '
'Can be absolute, or relative to the current working directory. '
'If not specified, SSL client authentication is disabled.')
parser.add_argument(
'--client-key',
help='path to a SSL client key. '
'Can be absolute, or relative to the current working directory. '
'Must be specified if "--client-cert" is provided.')
parser.add_argument('--basic-user',
help='username for basic authentication with nodes. '
'If not specified, basic authentication is disabled.')
parser.add_argument('--basic-password',
help='password for basic authentication with nodes. '
'Must be specified if "--basic-user" is provided.')
parser.add_argument(
'-p',
'--port',
type=int,
default=9206,
help='port to serve the metrics endpoint on. (default: 9206)')
parser.add_argument(
'--query-disable',
action='store_true',
help='disable query monitoring. '
'Config file does not need to be present if query monitoring is disabled.'
)
parser.add_argument(
'-c',
'--config-file',
default='exporter.cfg',
help='path to query config file. '
'Can be absolute, or relative to the current working directory. '
'(default: exporter.cfg)')
parser.add_argument(
'--config-dir',
default='./config',
help='path to query config directory. '
'Besides including the single config file specified by "--config-file" at first, '
'all config files in the config directory will be sorted, merged, then included. '
'Can be absolute, or relative to the current working directory. '
'(default: ./config)')
parser.add_argument('--cluster-health-disable',
action='store_true',
help='disable cluster health monitoring.')
parser.add_argument(
'--cluster-health-timeout',
type=float,
default=10.0,
help=
'request timeout for cluster health monitoring, in seconds. (default: 10)'
)
parser.add_argument(
'--cluster-health-level',
default='indices',
choices=['cluster', 'indices', 'shards'],
help=
'level of detail for cluster health monitoring. (default: indices)')
parser.add_argument('--nodes-stats-disable',
action='store_true',
help='disable nodes stats monitoring.')
parser.add_argument(
'--nodes-stats-timeout',
type=float,
default=10.0,
help=
'request timeout for nodes stats monitoring, in seconds. (default: 10)'
)
parser.add_argument(
'--nodes-stats-metrics',
type=nodes_stats_metrics_parser,
help='limit nodes stats to specific metrics. '
'Metrics should be separated by commas e.g. indices,fs.')
parser.add_argument('--indices-stats-disable',
action='store_true',
help='disable indices stats monitoring.')
parser.add_argument(
'--indices-stats-timeout',
type=float,
default=10.0,
help=
'request timeout for indices stats monitoring, in seconds. (default: 10)'
)
parser.add_argument(
'--indices-stats-mode',
default='cluster',
choices=['cluster', 'indices'],
help='detail mode for indices stats monitoring. (default: cluster)')
parser.add_argument(
'--indices-stats-metrics',
type=indices_stats_metrics_parser,
help='limit indices stats to specific metrics. '
'Metrics should be separated by commas e.g. indices,fs.')
parser.add_argument(
'--indices-stats-fields',
type=indices_stats_fields_parser,
help='include fielddata info for specific fields. '
'Fields should be separated by commas e.g. indices,fs. '
'Use \'*\' for all.')
parser.add_argument('-j',
'--json-logging',
action='store_true',
help='turn on json logging.')
parser.add_argument(
'--log-level',
default='INFO',
choices=['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'],
help='detail level to log. (default: INFO)')
parser.add_argument(
'-v',
'--verbose',
action='store_true',
help='turn on verbose (DEBUG) logging. Overrides --log-level.')
parser.add_argument('--indices-query-disable',
action='store_true',
help='disable indices query.')
args = parser.parse_args()
predefined_labels = OrderedDict()
if args.cluster_name is not None:
predefined_labels['cluster_name'] = [args.cluster_name]
if args.basic_user and args.basic_password is None:
parser.error('Username provided with no password.')
elif args.basic_user is None and args.basic_password:
parser.error('Password provided with no username.')
elif args.basic_user:
http_auth = (args.basic_user, args.basic_password)
else:
http_auth = None
if not args.ca_certs and (args.client_cert or args.client_key):
parser.error(
'--client-cert and --client-key can only be used when --ca-certs is provided.'
)
elif args.client_cert and not args.client_key:
parser.error(
'--client-key must be provided when --client-cert is used.')
elif not args.client_cert and args.client_key:
parser.error(
'--client-cert must be provided when --client-key is used.')
log_handler = logging.StreamHandler()
log_format = '[%(asctime)s] %(name)s.%(levelname)s %(threadName)s %(message)s'
formatter = JogFormatter(
log_format) if args.json_logging else logging.Formatter(log_format)
log_handler.setFormatter(formatter)
log_level = getattr(logging, args.log_level)
logging.basicConfig(handlers=[log_handler],
level=logging.DEBUG if args.verbose else log_level)
logging.captureWarnings(True)
port = args.port
es_cluster = args.es_cluster.split(',')
if args.ca_certs:
es_client = Elasticsearch(es_cluster,
verify_certs=True,
ca_certs=args.ca_certs,
client_cert=args.client_cert,
client_key=args.client_key,
http_auth=http_auth)
else:
es_client = Elasticsearch(es_cluster,
verify_certs=False,
http_auth=http_auth)
scheduler = None
print("query disable:", args.indices_query_disable)
if not args.indices_query_disable:
print("preparing query....")
scheduler = sched.scheduler()
config = configparser.ConfigParser()
config.read_file(open(args.config_file))
config_dir_file_pattern = os.path.join(args.config_dir, '*.cfg')
config_dir_sorted_files = sorted(glob.glob(config_dir_file_pattern))
config.read(config_dir_sorted_files)
query_prefix = 'query_'
queries = {}
for section in config.sections():
if section.startswith(query_prefix):
query_name = section[len(query_prefix):]
query_interval = config.getfloat(section,
'QueryIntervalSecs',
fallback=15)
query_timeout = config.getfloat(section,
'QueryTimeoutSecs',
fallback=10)
query_indices = config.get(section,
'QueryIndices',
fallback='_all')
query = json.loads(config.get(section, 'QueryJson'))
queries[query_name] = (query_interval, query_timeout,
query_indices, query)
if queries:
for name, (interval, timeout, indices, query) in queries.items():
func = partial(run_query, es_client, name, indices, query,
timeout, predefined_labels)
run_scheduler(scheduler, interval, func)
print("prepared query")
else:
logging.warn('No queries found in config file %s',
args.config_file)
if not args.cluster_health_disable:
REGISTRY.register(
ClusterHealthCollector(es_client,
args.cluster_health_timeout,
args.cluster_health_level,
predefined_labels=predefined_labels))
if not args.nodes_stats_disable:
REGISTRY.register(
NodesStatsCollector(es_client,
args.nodes_stats_timeout,
metrics=args.nodes_stats_metrics,
predefined_labels=predefined_labels))
if not args.indices_stats_disable:
parse_indices = args.indices_stats_mode == 'indices'
REGISTRY.register(
IndicesStatsCollector(es_client,
args.indices_stats_timeout,
parse_indices=parse_indices,
metrics=args.indices_stats_metrics,
fields=args.indices_stats_fields,
predefined_labels=predefined_labels))
logging.info('Starting server...')
start_http_server(port)
logging.info('Server started on port %s', port)
try:
if scheduler:
scheduler.run()
else:
while True:
time.sleep(5)
except KeyboardInterrupt:
pass
shutdown()
def sanitise_name(s):
return re.sub(r"[^a-zA-Z0-9:_]", "_", s)
class ConsulCollector(object):
def collect(self):
out = urlopen("http://localhost:8500/v1/agent/metrics").read()
metrics = json.loads(out.decode("utf-8"))
for g in metrics["Gauges"]:
yield GaugeMetricFamily(sanitise_name(g["Name"]),
"Consul metric " + g["Name"], g["Value"])
for c in metrics["Counters"]:
yield CounterMetricFamily(sanitise_name(c["Name"]) + "_total",
"Consul metric " + c["Name"], c["Count"])
for s in metrics["Samples"]:
yield SummaryMetricFamily(sanitise_name(s["Name"]) + "_seconds",
"Consul metric " + s["Name"],
count_value=c["Count"], sum_value=s["Sum"] / 1000)
if __name__ == '__main__':
REGISTRY.register(ConsulCollector())
start_http_server(8000)
while True:
time.sleep(1)
for x in range(get_db['replic_status']):
replic_usesysid.add_metric([get_db['db_name'], get_db['replic_ip'][x]], get_db['replic_usesysid'][x])
replic_pid.add_metric([get_db['db_name'], get_db['replic_ip'][x]], get_db['replic_pid'][x])
replica_lags.add_metric([get_db['db_name'], get_db['replic_ip'][x]], get_db['replica_lags'][x])
yield size
yield max_connections
yield total_connections
yield left_connections
yield db_deadlocks
yield replica_lags
yield replic_usesysid
yield replic_pid
yield replic_status
if __name__ == '__main__':
# Start up the server to expose the metrics.
start_http_server(server_port)
# Generate some requests.
get_db = postgres(host, dbname, user, password, dbname_postgres)
REGISTRY.register(CustomCollector())
while True:
time.sleep(1)
get_db = postgres(host, dbname, user, password, dbname_postgres)
# while True: time.sleep(1)
def main():
REGISTRY.register(IpmiCollector())
start_http_server(8000)
while True:
time.sleep(5)
parser.add_argument('--secure', default='no', type=str)
args = parser.parse_args()
# Wait for other containers to start.
print('>>> Sleeping for: %s seconds ...' % args.start_delay)
time.sleep(args.start_delay)
# Start the server to expose the metrics.
print('>>> Starting the exporter on port: %s' % args.port)
start_http_server(args.port)
# Get username and password of NetScalers.
ns_user = os.environ.get("NS_USER")
ns_password = os.environ.get("NS_PASSWORD")
if ns_user == None:
ns_user = args.username
if ns_password == None:
ns_password = args.password
# Load the metrics file specifying stats to be collected
f = open('/exporter/metrics.json', 'r')
metrics_json = json.load(f)
# Register the exporter as a stat collector.
print('>>> Registering collector for: %s' % (args.target_nsip))
REGISTRY.register(NetscalerCollector(nsips=args.target_nsip, metrics=metrics_json, username=ns_user, password=ns_password, secure=args.secure.lower()))
# Forever
while True:
time.sleep(1)
def start():
REGISTRY.register(
CollectorHandler(RouterEntriesHandler(), CollectorRegistry()))
ExportProcessor.run(port=config_handler.system_entry().port)
def main():
signal.signal(signal.SIGTERM, signal_handler)
parser = argparse.ArgumentParser(
description='Export Kafka consumer offsets to Prometheus.')
parser.add_argument(
'-b',
'--bootstrap-brokers',
default='localhost',
help='Addresses of brokers in a Kafka cluster to talk to.' +
' Brokers should be separated by commas e.g. broker1,broker2.' +
' Ports can be provided if non-standard (9092) e.g. brokers1:9999.' +
' (default: localhost)')
parser.add_argument(
'-p',
'--port',
type=int,
default=9208,
help='Port to serve the metrics endpoint on. (default: 9208)')
parser.add_argument(
'-s',
'--from-start',
action='store_true',
help='Start from the beginning of the `__consumer_offsets` topic.')
parser.add_argument(
'--topic-interval',
type=float,
default=30.0,
help='How often to refresh topic information, in seconds. (default: 30)'
)
parser.add_argument(
'--high-water-interval',
type=float,
default=10.0,
help=
'How often to refresh high-water information, in seconds. (default: 10)'
)
parser.add_argument(
'--low-water-interval',
type=float,
default=10.0,
help=
'How often to refresh low-water information, in seconds. (default: 10)'
)
parser.add_argument(
'--consumer-config',
action='append',
default=[],
help=
'Provide additional Kafka consumer config as a consumer.properties file. Multiple files will be merged, later files having precedence.'
)
parser.add_argument('-j',
'--json-logging',
action='store_true',
help='Turn on json logging.')
parser.add_argument(
'--log-level',
default='INFO',
choices=['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'],
help='detail level to log. (default: INFO)')
parser.add_argument(
'-v',
'--verbose',
action='store_true',
help='turn on verbose (DEBUG) logging. Overrides --log-level.')
args = parser.parse_args()
log_handler = logging.StreamHandler()
log_format = '[%(asctime)s] %(name)s.%(levelname)s %(threadName)s %(message)s'
formatter = JogFormatter(log_format) \
if args.json_logging \
else logging.Formatter(log_format)
log_handler.setFormatter(formatter)
log_level = getattr(logging, args.log_level)
logging.basicConfig(handlers=[log_handler],
level=logging.DEBUG if args.verbose else log_level)
logging.captureWarnings(True)
port = args.port
consumer_config = {
'bootstrap_servers': 'localhost',
'auto_offset_reset': 'latest',
'group_id': None,
'consumer_timeout_ms': 500
}
for filename in args.consumer_config:
with open(filename) as f:
raw_config = javaproperties.load(f)
converted_config = {
k.replace('.', '_'): v
for k, v in raw_config.items()
}
consumer_config.update(converted_config)
if args.bootstrap_brokers:
consumer_config['bootstrap_servers'] = args.bootstrap_brokers.split(
',')
if args.from_start:
consumer_config['auto_offset_reset'] = 'earliest'
consumer = KafkaConsumer('__consumer_offsets', **consumer_config)
client = consumer._client
topic_interval = args.topic_interval
high_water_interval = args.high_water_interval
low_water_interval = args.low_water_interval
logging.info('Starting server...')
start_http_server(port)
logging.info('Server started on port %s', port)
REGISTRY.register(collectors.HighwaterCollector())
REGISTRY.register(collectors.LowwaterCollector())
REGISTRY.register(collectors.ConsumerOffsetCollector())
REGISTRY.register(collectors.ConsumerLagCollector())
REGISTRY.register(collectors.ConsumerLeadCollector())
REGISTRY.register(collectors.ConsumerCommitsCollector())
REGISTRY.register(collectors.ExporterOffsetCollector())
REGISTRY.register(collectors.ExporterLagCollector())
REGISTRY.register(collectors.ExporterLeadCollector())
scheduled_jobs = setup_fetch_jobs(topic_interval, high_water_interval,
low_water_interval, client)
try:
while True:
for message in consumer:
offsets = collectors.get_offsets()
commits = collectors.get_commits()
exporter_offsets = collectors.get_exporter_offsets()
exporter_partition = message.partition
exporter_offset = message.offset
exporter_offsets = ensure_dict_key(exporter_offsets,
exporter_partition,
exporter_offset)
exporter_offsets[exporter_partition] = exporter_offset
collectors.set_exporter_offsets(exporter_offsets)
if message.key and message.value:
key = parse_key(message.key)
if key:
value = parse_value(message.value)
group = key[1]
topic = key[2]
partition = key[3]
offset = value[1]
offsets = ensure_dict_key(offsets, group, {})
offsets[group] = ensure_dict_key(
offsets[group], topic, {})
offsets[group][topic] = ensure_dict_key(
offsets[group][topic], partition, offset)
offsets[group][topic][partition] = offset
collectors.set_offsets(offsets)
commits = ensure_dict_key(commits, group, {})
commits[group] = ensure_dict_key(
commits[group], topic, {})
commits[group][topic] = ensure_dict_key(
commits[group][topic], partition, 0)
commits[group][topic][partition] += 1
collectors.set_commits(commits)
# Check if we need to run any scheduled jobs
# each message.
scheduled_jobs = scheduler.run_scheduled_jobs(scheduled_jobs)
# Also check if we need to run any scheduled jobs
# each time the consumer times out, in case there
# aren't any messages to consume.
scheduled_jobs = scheduler.run_scheduled_jobs(scheduled_jobs)
except KeyboardInterrupt:
pass
shutdown()
def collect():
for metric in REGISTRY.collect():
if not metric.name.startswith("__"):
yield metric
bom_pressure_pascals.add_metric(labels, latest_obs["press"] *
100) # hPa to Pa
if latest_obs["rel_hum"] != None:
bom_relative_humidity.add_metric(labels, latest_obs["rel_hum"])
if latest_obs["wind_spd_kmh"] != None:
bom_wind_speed.add_metric(labels, latest_obs["wind_spd_kmh"])
if latest_obs["apparent_t"] != None:
bom_apparent_temperature_celsius.add_metric(
labels, latest_obs["apparent_t"])
wind_dir = latest_obs["wind_dir"]
if wind_dir in WIND_DIR_TO_DEGREES:
bom_wind_direction_degrees.add_metric(
labels, WIND_DIR_TO_DEGREES[wind_dir])
yield bom_utctimestamp
yield bom_air_temperature
yield bom_pressure_pascals
yield bom_relative_humidity
yield bom_wind_speed
yield bom_wind_direction_degrees
yield bom_apparent_temperature_celsius
# Query the BOM
REGISTRY.register(BOMCollector())
start_http_server(8000)
while True:
time.sleep(60)
metric.build_name = build.metadata.name
metric.build_config_name = build.metadata.labels.buildconfig
metric.namespace = build.metadata.namespace
labels = build.metadata.labels
metric.labels = json.loads(str(labels).replace("\'", "\""))
metric.commit_hash = build.spec.revision.git.commit
metric.name = app + '-' + build.spec.revision.git.commit
metric.commiter = build.spec.revision.git.author.name
metric.image_location = build.status.outputDockerImageReference
metric.image_hash = build.status.output.to.imageDigest
metric.getCommitTime()
metrics.append(metric)
return metrics
if __name__ == "__main__":
username = os.environ.get('GITHUB_USER')
token = os.environ.get('GITHUB_TOKEN')
namespaces = None
if os.environ.get('NAMESPACES') is not None:
namespaces = [
proj.strip() for proj in os.environ.get('NAMESPACES').split(",")
]
apps = None
start_http_server(8080)
REGISTRY.register(CommitCollector(username, token, namespaces, apps))
while True:
time.sleep(1)
def register_cache(cache_type, cache_name, cache, collect_callback=None):
"""Register a cache object for metric collection.
Args:
cache_type (str):
cache_name (str): name of the cache
cache (object): cache itself
collect_callback (callable|None): if not None, a function which is called during
metric collection to update additional metrics.
Returns:
CacheMetric: an object which provides inc_{hits,misses,evictions} methods
"""
# Check if the metric is already registered. Unregister it, if so.
# This usually happens during tests, as at runtime these caches are
# effectively singletons.
metric_name = "cache_%s_%s" % (cache_type, cache_name)
if metric_name in collectors_by_name.keys():
REGISTRY.unregister(collectors_by_name[metric_name])
class CacheMetric(object):
hits = 0
misses = 0
evicted_size = 0
def inc_hits(self):
self.hits += 1
def inc_misses(self):
self.misses += 1
def inc_evictions(self, size=1):
self.evicted_size += size
def describe(self):
return []
def collect(self):
try:
if cache_type == "response_cache":
response_cache_size.labels(cache_name).set(len(cache))
response_cache_hits.labels(cache_name).set(self.hits)
response_cache_evicted.labels(cache_name).set(
self.evicted_size)
response_cache_total.labels(cache_name).set(self.hits +
self.misses)
else:
cache_size.labels(cache_name).set(len(cache))
cache_hits.labels(cache_name).set(self.hits)
cache_evicted.labels(cache_name).set(self.evicted_size)
cache_total.labels(cache_name).set(self.hits + self.misses)
if collect_callback:
collect_callback()
except Exception as e:
logger.warn("Error calculating metrics for %s: %s", cache_name,
e)
raise
yield GaugeMetricFamily("__unused", "")
metric = CacheMetric()
REGISTRY.register(metric)
caches_by_name[cache_name] = cache
collectors_by_name[metric_name] = metric
return metric
return getattr(self._poller, item)
class ReactorLastSeenMetric(Collector):
def __init__(self, epoll_wrapper: EpollWrapper):
self._epoll_wrapper = epoll_wrapper
def collect(self) -> Iterable[Metric]:
cm = GaugeMetricFamily(
"python_twisted_reactor_last_seen",
"Seconds since the Twisted reactor was last seen",
)
cm.add_metric([], time.time() - self._epoll_wrapper.last_polled)
yield cm
try:
# if the reactor has a `_poller` attribute, which is an `epoll` object
# (ie, it's an EPollReactor), we wrap the `epoll` with a thing that will
# measure the time between ticks
from select import epoll # type: ignore[attr-defined]
poller = reactor._poller # type: ignore[attr-defined]
except (AttributeError, ImportError):
pass
else:
if isinstance(poller, epoll):
poller = EpollWrapper(poller)
reactor._poller = poller # type: ignore[attr-defined]
REGISTRY.register(ReactorLastSeenMetric(poller))
conn.close()
def assign_location(self):
for jail in self.jails:
for entry in jail.ip_list:
entry.update(self.geo_provider.annotate(entry['ip']))
def collect(self):
self.get_jailed_ips()
self.assign_location()
metric_labels = ['jail','ip'] + self.extra_labels
ip_gauge = GaugeMetricFamily('fail2ban_banned_ip', 'IP banned by fail2ban', labels=metric_labels)
for jail in self.jails:
for entry in jail.ip_list:
values = [jail.name, entry['ip']] + [entry[x] for x in self.extra_labels]
ip_gauge.add_metric(values, 1)
yield ip_gauge
if __name__ == '__main__':
with open('conf.yml') as f:
conf = yaml.load(f, Loader=yaml.FullLoader)
REGISTRY.register(F2bCollector(conf))
app = make_wsgi_app()
httpd = make_server('', conf['server']['port'], app)
httpd.serve_forever()
# Verify ADC stats access
verify_ns_stats_access(args.target_nsip, ns_protocol, ns_user, ns_password,
args.timeout)
# Start the server to expose the metrics.
start_exporter_server(args.port)
if not args.k8sCICprefix.isalnum():
logger.error('Invalid k8sCICprefix : non-alphanumeric not accepted')
# Register the exporter as a stat collector
logger.info('Registering collector for %s' % args.target_nsip)
try:
REGISTRY.register(
CitrixAdcCollector(nsip=args.target_nsip,
metrics=metrics_json,
username=ns_user,
password=ns_password,
protocol=ns_protocol,
nitro_timeout=args.timeout,
k8s_cic_prefix=args.k8sCICprefix))
except Exception as e:
logger.error(
'Invalid arguments! could not register collector for {}::{}'.
format(args.target_nsip, e))
# Forever
while True:
signal.pause()
]
try:
LINKY_EXPORTER_PORT = int(os.environ.get('LINKY_EXPORTER_PORT', '8123'))
except ValueError:
logging.error("LINKY_EXPORTER_PORT must be int !")
sys.exit(1)
LINKY_EXPORTER_MODE = os.environ.get('LINKY_EXPORTER_MODE', 'HISTORIQUE')
VALID_MODE = [i['name'] for i in LINKY_MODE]
if not LINKY_EXPORTER_MODE in VALID_MODE:
logging.error("LINKY_EXPORTER_MODE must be : %s", ' or '.join(VALID_MODE))
sys.exit(1)
# REGISTRY Configuration
REGISTRY.unregister(PROCESS_COLLECTOR)
REGISTRY.unregister(PLATFORM_COLLECTOR)
REGISTRY.unregister(REGISTRY._names_to_collectors['python_gc_objects_collected_total'])
# Linky Collector Class
class LinkyCollector():
'''Linky Collector Class'''
def __init__(self):
self.ser = self._check_for_valid_frame()
def teleinfo(self):
'''Read Teleinfo And Return Linky Frame Dict'''
logging.debug("Reading Linky Teleinfo on %s.", LINKY_EXPORTER_INTERFACE)
with self.ser:
# Wait For New Linky Frame (Start with 0x02)
parser.add_argument("-s",
"--control-socket",
help="Tor control socket",
default="/var/run/tor/control")
parser.add_argument("-p",
"--listen-port",
help="Listen on this port",
type=int,
default=9099)
parser.add_argument("-b",
"--bind-addr",
help="Bind this address",
default="localhost")
args = parser.parse_args()
if args.mode == 'unix':
torctl = stem.control.Controller.from_socket_file(args.control_socket)
else:
torctl = stem.control.Controller.from_port(args.address,
port=args.control_port)
coll = StemCollector(torctl)
REGISTRY.register(coll)
print("Starting on %s:%s" % (args.bind_addr, args.listen_port))
prom.start_http_server(args.listen_port, addr=args.bind_addr)
# We can't exit as start_http_server starts a daemon thread which would get
# killed.
while True:
time.sleep(1000)
yield background_process_in_flight_count
# now we need to run collect() over each of the static Counters, and
# yield each metric they return.
for m in (
_background_process_ru_utime,
_background_process_ru_stime,
_background_process_db_txn_count,
_background_process_db_txn_duration,
_background_process_db_sched_duration,
):
for r in m.collect():
yield r
REGISTRY.register(_Collector())
class _BackgroundProcess(object):
def __init__(self, desc, ctx):
self.desc = desc
self._context = ctx
self._reported_stats = None
def update_metrics(self):
"""Updates the metrics with values from this process."""
new_stats = self._context.get_resource_usage()
if self._reported_stats is None:
diff = new_stats
else:
diff = new_stats - self._reported_stats
def _collect_to_http():
REGISTRY.register(BitfinexCollector())
start_http_server(int(settings['bitfinex_exporter']['listen_port']))
while True:
time.sleep(int(settings['bitfinex_exporter']['interval']))
process.update_metrics()
# now we need to run collect() over each of the static Counters, and
# yield each metric they return.
for m in (
_background_process_ru_utime,
_background_process_ru_stime,
_background_process_db_txn_count,
_background_process_db_txn_duration,
_background_process_db_sched_duration,
):
for r in m.collect():
yield r
REGISTRY.register(_Collector())
class _BackgroundProcess:
def __init__(self, desc, ctx):
self.desc = desc
self._context = ctx
self._reported_stats = None
def update_metrics(self):
"""Updates the metrics with values from this process."""
new_stats = self._context.get_resource_usage()
if self._reported_stats is None:
diff = new_stats
else:
diff = new_stats - self._reported_stats
def collect(self):
temp_guage = GaugeMetricFamily('temperature',
'Current temperature in celsius',
labels=['host', 'sensor_type'])
temp_guage.add_metric([self.hostname, 'BME280'],
round(self.sensor.temperature, 2))
yield temp_guage
humidity_guage = GaugeMetricFamily('humidity',
'Current realtive humidity',
labels=['host', 'sensor_type'])
humidity_guage.add_metric([self.hostname, 'BME280'],
round(self.sensor.humidity, 2))
yield humidity_guage
pressure_guage = GaugeMetricFamily(
'pressure',
'Current atmospheric pressure in hPa',
labels=['host', 'sensor_type'])
pressure_guage.add_metric([self.hostname, 'BME280'],
round(self.sensor.pressure, 2))
yield pressure_guage
if __name__ == "__main__":
REGISTRY.register(BMP280Exporter())
start_http_server(8000)
running = True
while running:
try:
time.sleep(10)
except KeyboardInterrupt:
running = False
"Syncing": 3,
"Farming": 4
}
def __init__(self, target):
self.target = target
def get(self):
response = requests.get(self.target)
return response.json()
def collect(self):
data = self.get()
for name, value in data.items():
if name == "status":
value = self.farm_status[value]
yield GaugeMetricFamily(f"chia_farm_summary_{name}",
f"chia_farm_summary_{name}",
value=value)
#stdout = subprocess.check_output(["plotman", "status"]).decode("utf-8")
#yield GaugeMetricFamily("chia_farm_summary_plots_in_progress", "plotman_jobs_count", value=stdout.count("/mnt/plotter"))
if __name__ == "__main__":
port = 8000
start_http_server(port)
REGISTRY.register(ChiaCollector(CHIA_HTTP_SERVER))
print(f"Listening on port {port}")
while True:
time.sleep(1)
return GitLabCommitCollector("", "", "", "")
if git_provider == "github":
return GitHubCommitCollector(username, token, namespaces, apps,
git_api)
if git_provider == "bitbucket":
return BitbucketCommitCollector("", "", "", "")
if __name__ == "__main__":
pelorus.check_legacy_vars()
pelorus.check_required_config(REQUIRED_CONFIG)
username = os.environ.get('GIT_USER')
token = os.environ.get('GIT_TOKEN')
git_api = os.environ.get('GIT_API')
git_provider = os.environ.get('GIT_PROVIDER', pelorus.DEFAULT_GIT)
namespaces = None
if os.environ.get('NAMESPACES') is not None:
namespaces = [
proj.strip() for proj in os.environ.get('NAMESPACES').split(",")
]
apps = None
start_http_server(8080)
collector = GitFactory.getCollector(username, token, namespaces, apps,
git_api, git_provider)
REGISTRY.register(collector)
while True:
time.sleep(1)
import os
import logging
from flask import Flask, redirect, Response
from collector import MarathonAppCollector
from prometheus_client import PROCESS_COLLECTOR
from prometheus_client.core import REGISTRY
from prometheus_client.exposition import generate_latest
MARATHON_URL = os.environ.get(
'MARATHON_URL',
'http://leader.mesos:8080/')
REGISTRY.unregister(PROCESS_COLLECTOR)
REGISTRY.register(MarathonAppCollector(MARATHON_URL))
app = Flask(__name__)
@app.route('/')
def home():
return redirect('/metrics')
@app.route('/metrics')
def metrics():
prom_metrics = generate_latest(REGISTRY)
return Response(prom_metrics, content_type='text/plain')
if __name__ == '__main__':
log_format = u'[%(asctime)s] p%(process)s {%(pathname)s:%(lineno)d}' \
u' %(levelname)s - %(message)s'
logging.basicConfig(
class GaugeCollector(object):
def collect(self):
for jail in self.get_jails(self.extract_data()):
g = GaugeMetricFamily("fail2ban_{}".format(self.snake_case(jail)),
"",
labels=['type'])
for label, value in self.extract_data(jail):
g.add_metric([self.snake_case(label)], float(value))
yield g
def get_jails(self, jails):
return jails[1][1].split(",")
def extract_data(self, jail=""):
r = run(cmd.format(path=path, service=jail),
stdout=PIPE,
check=True,
shell=True)
return findall(comp, ''.join(bytes(r.stdout).decode('utf-8')).lower())
def snake_case(self, string):
return string.strip().replace("-", "_").replace(" ", "_")
# Code execution starts from here
start_http_server(port, addr)
REGISTRY.register(GaugeCollector())
while True:
sleep(10)
return
with open("/proc/self/stat") as s:
line = s.read()
raw_stats = line.split(") ", 1)[1].split(" ")
user = GaugeMetricFamily("process_cpu_user_seconds_total", "")
user.add_metric([], float(raw_stats[11]) / self.ticks_per_sec)
yield user
sys = GaugeMetricFamily("process_cpu_system_seconds_total", "")
sys.add_metric([], float(raw_stats[12]) / self.ticks_per_sec)
yield sys
REGISTRY.register(CPUMetrics())
#
# Python GC metrics
#
gc_unreachable = Gauge("python_gc_unreachable_total", "Unreachable GC objects", ["gen"])
gc_time = Histogram(
"python_gc_time",
"Time taken to GC (sec)",
["gen"],
buckets=[0.0025, 0.005, 0.01, 0.025, 0.05, 0.10, 0.25, 0.50, 1.00, 2.50,
5.00, 7.50, 15.00, 30.00, 45.00, 60.00],
)
def cli(**options):
"""Export Elasticsearch query results to Prometheus."""
if options['basic_user'] and options['basic_password'] is None:
raise click.BadOptionUsage('basic_user',
'Username provided with no password.')
elif options['basic_user'] is None and options['basic_password']:
raise click.BadOptionUsage('basic_password',
'Password provided with no username.')
elif options['basic_user']:
http_auth = (options['basic_user'], options['basic_password'])
else:
http_auth = None
if not options['ca_certs'] and options['client_cert']:
raise click.BadOptionUsage(
'client_cert',
'--client-cert can only be used when --ca-certs is provided.')
elif not options['ca_certs'] and options['client_key']:
raise click.BadOptionUsage(
'client_key',
'--client-key can only be used when --ca-certs is provided.')
elif options['client_cert'] and not options['client_key']:
raise click.BadOptionUsage(
'client_cert',
'--client-key must be provided when --client-cert is used.')
elif not options['client_cert'] and options['client_key']:
raise click.BadOptionUsage(
'client_key',
'--client-cert must be provided when --client-key is used.')
if options['indices_stats_indices'] and options[
'indices_stats_mode'] != 'indices':
raise click.BadOptionUsage(
'indices_stats_indices',
'--indices-stats-mode must be "indices" for '
'--indices-stats-indices to be used.')
log_handler = logging.StreamHandler()
log_format = '[%(asctime)s] %(name)s.%(levelname)s %(threadName)s %(message)s'
formatter = JogFormatter(
log_format) if options['json_logging'] else logging.Formatter(
log_format)
log_handler.setFormatter(formatter)
log_level = getattr(logging, options['log_level'])
logging.basicConfig(
handlers=[log_handler],
level=logging.DEBUG if options['verbose'] else log_level)
logging.captureWarnings(True)
port = options['port']
es_cluster = options['es_cluster'].split(',')
if options['ca_certs']:
es_client = Elasticsearch(es_cluster,
verify_certs=True,
ca_certs=options['ca_certs'],
client_cert=options['client_cert'],
client_key=options['client_key'],
headers=options['header'],
http_auth=http_auth)
else:
es_client = Elasticsearch(es_cluster,
verify_certs=False,
headers=options['header'],
http_auth=http_auth)
scheduler = None
if not options['query_disable']:
config = configparser.ConfigParser(converters=CONFIGPARSER_CONVERTERS)
config.read(options['config_file'])
config_dir_file_pattern = os.path.join(options['config_dir'], '*.cfg')
config_dir_sorted_files = sorted(glob.glob(config_dir_file_pattern))
config.read(config_dir_sorted_files)
query_prefix = 'query_'
queries = {}
for section in config.sections():
if section.startswith(query_prefix):
query_name = section[len(query_prefix):]
interval = config.getfloat(section,
'QueryIntervalSecs',
fallback=15)
timeout = config.getfloat(section,
'QueryTimeoutSecs',
fallback=10)
indices = config.get(section, 'QueryIndices', fallback='_all')
query = json.loads(config.get(section, 'QueryJson'))
on_error = config.getenum(section,
'QueryOnError',
fallback='drop')
on_missing = config.getenum(section,
'QueryOnMissing',
fallback='drop')
queries[query_name] = (interval, timeout, indices, query,
on_error, on_missing)
scheduler = sched.scheduler()
if queries:
for query_name, (interval, timeout, indices, query, on_error,
on_missing) in queries.items():
schedule_job(scheduler, interval, run_query, es_client,
query_name, indices, query, timeout, on_error,
on_missing)
else:
log.error('No queries found in config file(s)')
return
if not options['cluster_health_disable']:
REGISTRY.register(
ClusterHealthCollector(es_client,
options['cluster_health_timeout'],
options['cluster_health_level']))
if not options['nodes_stats_disable']:
REGISTRY.register(
NodesStatsCollector(es_client,
options['nodes_stats_timeout'],
metrics=options['nodes_stats_metrics']))
if not options['indices_aliases_disable']:
REGISTRY.register(
IndicesAliasesCollector(es_client,
options['indices_aliases_timeout']))
if not options['indices_mappings_disable']:
REGISTRY.register(
IndicesMappingsCollector(es_client,
options['indices_mappings_timeout']))
if not options['indices_stats_disable']:
parse_indices = options['indices_stats_mode'] == 'indices'
REGISTRY.register(
IndicesStatsCollector(es_client,
options['indices_stats_timeout'],
parse_indices=parse_indices,
indices=options['indices_stats_indices'],
metrics=options['indices_stats_metrics'],
fields=options['indices_stats_fields']))
if scheduler:
REGISTRY.register(QueryMetricCollector())
log.info('Starting server...')
start_http_server(port)
log.info('Server started on port %(port)s', {'port': port})
if scheduler:
scheduler.run()
else:
while True:
time.sleep(5)
)
class Expositor(object):
""" Responsible for exposing metrics to prometheus """
def collect(self):
logging.info("Serving prometheus data")
for key in sorted(metrics):
yield metrics[key]
if __name__ == '__main__':
logging.basicConfig(level=logging.DEBUG)
for collector in list(REGISTRY._collector_to_names):
REGISTRY.unregister(collector)
REGISTRY.register(Expositor())
start_time = time.time()
# Popluate data before exposing over http
scrape()
start_http_server(8000)
ready_time = time.time()
print("Ready time: ", ready_time-start_time)
while True:
time.sleep(int(os.environ.get('KOJI_POLL_INTERVAL', '3')))
scrape()
finally:
response_json.append(data)
#Generate metric output
for item in range(len(response_json)):
docker_namespace = response_json[item].get('namespace')
docker_name = response_json[item].get('name')
docker_namespace = docker_namespace.translate(table)
docker_name = docker_name.translate(table)
g = GaugeMetricFamily(
'dockerpulls_' + docker_namespace + '_' + docker_name +
'_total', 'Total Pulls for: ' +
response_json[item].get('namespace') + '/' +
response_json[item].get('name'))
g.add_metric('docker_pulls',
response_json[item].get('pull_count'))
yield g
#Register the collector for metrics display
REGISTRY.register(CustomCollector())
#Sets the /metrics
#/ has a welcoming message
def docker_pull(environ, start_fn):
if environ['PATH_INFO'] == '/metrics':
return metrics_app(environ, start_fn)
start_fn('200 OK', [])
return [b'Hi there \\o\n\nMaybe you wanna go to /metrics!? :)']
return
with open("/proc/self/stat") as s:
line = s.read()
raw_stats = line.split(") ", 1)[1].split(" ")
user = GaugeMetricFamily("process_cpu_user_seconds_total", "")
user.add_metric([], float(raw_stats[11]) / self.ticks_per_sec)
yield user
sys = GaugeMetricFamily("process_cpu_system_seconds_total", "")
sys.add_metric([], float(raw_stats[12]) / self.ticks_per_sec)
yield sys
REGISTRY.register(CPUMetrics())
#
# Python GC metrics
#
gc_unreachable = Gauge("python_gc_unreachable_total", "Unreachable GC objects",
["gen"])
gc_time = Histogram(
"python_gc_time",
"Time taken to GC (sec)",
["gen"],
buckets=[
0.0025,
0.005,
0.01,
def collect():
for metric in REGISTRY.collect():
if not metric.name.startswith("__"):
yield metric
parser.add_argument('--bamboo_url', type=str, required=True, env_var='bamboo_url')
parser.add_argument('--bamboo_user', type=str, required=True, env_var='bamboo_user')
parser.add_argument('--bamboo_password', type=str, required=True, env_var='bamboo_password')
parser.add_argument('--bamboo_test_jobs', type=str, env_var='bamboo_test_jobs') # CSV of PRJ-XX-JOB, eg COS-COS1-DTOL
# parser.add_argument('--dashing_event_url', type=str, required=True, env_var='dashing_event_url')
# parser.add_argument('--redis_host', type=str, required=True, env_var='redis_host')
# parser.add_argument('--redis_port', type=int, env_var='redis_port', default=6379)
parser.add_argument('--run_once', action='store_true')
parser.add_argument('--sonar_url', type=str, required=True, env_var='sonar_url')
parser.add_argument('--sonar_user', type=str, required=True, env_var='sonar_user')
parser.add_argument('--sonar_password', type=str, required=True, env_var='sonar_password')
args = parser.parse_args()
REGISTRY.register(BambooCollector(args.bamboo_url, args.bamboo_user, args.bamboo_password, args.bamboo_test_jobs))
# REGISTRY.register(EventStreamCollector(args.dashing_event_url)) # http://192.168.99.100:3030/events
# REGISTRY.register(RedisCollector(args.redis_host, args.redis_port))
REGISTRY.register(SonarCollector(args.sonar_url, args.sonar_user, args.sonar_password, []))
if args.run_once:
# time.sleep(5) # wait for async
pp = pprint.PrettyPrinter(indent=4)
for collector in REGISTRY._collectors:
# collector = BambooCollector(args.bamboo_url, args.bamboo_user, args.bamboo_password)
print collector
for x in collector.collect():
pp.pprint(x.samples)
sys.exit("runonce")
start_http_server(9118)
import configparser
from prometheus_client import make_wsgi_app
from prometheus_client.core import REGISTRY
from flask import Flask
from werkzeug.middleware.dispatcher import DispatcherMiddleware
from sonar.sonar import SonarCollector
# Import configuration file
config = configparser.ConfigParser()
config.read("config.ini")
sonar_collector = SonarCollector(
server=config['DEFAULT']['SONAR_SERVER'],
user=config['DEFAULT']['SONAR_USERNAME'],
passwd=config['DEFAULT']['SONAR_PASSWORD']
)
REGISTRY.register(sonar_collector)
# Create Flask app
app = Flask(__name__)
@app.route('/ready')
def ready():
return 'To infinity and beyond!'
# Add prometheus wsgi middleware to route /metrics requests
app.wsgi_app = DispatcherMiddleware(app.wsgi_app, {
'/metrics': make_wsgi_app()
})
'Jenkins build timestamp in unixtime for {0}'.format(s), labels=["jobname"]),
}
# Request exactly the information we need from Jenkins
result = json.loads(urllib2.urlopen(
"{0}/api/json?tree=jobs[name,{1}]".format(
self._target, ",".join([s + "[number,timestamp,duration]" for s in statuses])))
.read().decode("utf-8"))
for job in result['jobs']:
name = job['name']
for s in statuses:
# If there's a null result, we want to export zeros.
status = job[s] or {}
metrics[s]['number'].add_metric([name], status.get('number', 0))
metrics[s]['duration'].add_metric([name], status.get('duration', 0) / 1000.0)
metrics[s]['timestamp'].add_metric([name], status.get('timestamp', 0) / 1000.0)
for s in statuses:
for m in metrics[s].values():
yield m
if __name__ == "__main__":
if len(sys.argv) < 2:
sys.stderr.write("Usage: jenkins_exporter.py http://jenkins:8080\n")
sys.exit(1)
REGISTRY.register(JenkinsCollector(sys.argv[1]))
start_http_server(9118)
while True: time.sleep(1)