monasca-agent/monasca_agent/collector/checks_d/cadvisor_host.py

274 lines
13 KiB
Python

# (C) Copyright 2017 Hewlett Packard Enterprise Development LP
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import requests
from six.moves.urllib.parse import urlparse
from six.moves.urllib.parse import urlunparse
from monasca_agent.collector.checks import AgentCheck
from monasca_agent.collector.checks import utils
from monasca_agent.common.util import rollup_dictionaries
DEFAULT_TIMEOUT = "3"
# format for METRICS: (cadvisor metric name, [metric types], [metric units])
METRICS = {
"cpu_metrics": {
"system": ("system_time", ["gauge", "rate"],
["core_seconds", "core_seconds_per_second"]),
"total": ("total_time", ["gauge", "rate"],
["core_seconds", "core_seconds_per_second"]),
"user": ("user_time", ["gauge", "rate"],
["core_seconds", "core_seconds_per_second"])
},
"memory_metrics": {
"swap": ("swap_bytes", ["gauge"], ["bytes"]),
"cache": ("cache_bytes", ["gauge"], ["bytes"]),
"usage": ("used_bytes", ["gauge"], ["bytes"]),
"working_set": ("working_set", ["gauge"], ["bytes"])
},
"filesystem_metrics": {
"capacity": ("total_bytes", ["gauge"], ["bytes"]),
"usage": ("usage_bytes", ["gauge"], ["bytes"])
},
'network_metrics': {
"rx_bytes": ("in_bytes", ["gauge", "rate"],
["total_bytes", "total_bytes_per_second"]),
"tx_bytes": ("out_bytes", ["gauge", "rate"],
["total_bytes", "total_bytes_per_second"]),
"rx_packets": ("in_packets", ["gauge", "rate"],
["total_packets", "total_packets_per_second"]),
"tx_packets": ("out_packets", ["gauge", "rate"],
["total_packets", "total_packets_per_second"]),
"rx_dropped": ("in_dropped_packets", ["gauge", "rate"],
["total_packets", "total_packets_per_second"]),
"tx_dropped": ("out_dropped_packets", ["gauge", "rate"],
["total_packets", "total_packets_per_second"]),
"rx_errors": ("in_errors", ["gauge", "rate"],
["total_errors", "total_errors_per_second"]),
"tx_errors": ("out_errors", ["gauge", "rate"],
["total_errors", "total_errors_per_second"])
},
"diskio_metrics": {
"Read": ("read_bytes", ["gauge", "rate"],
["total_bytes", "total_bytes_per_second"]),
'Write': ("write_bytes", ["gauge", "rate"],
["total_bytes", "total_bytes_per_second"])
},
}
class CadvisorHost(AgentCheck):
"""Queries given cAdvisor API for node metrics
"""
def __init__(self, name, init_config, agent_config, instances=None):
AgentCheck.__init__(self, name, init_config, agent_config, instances)
if instances is not None and len(instances) > 1:
raise Exception('cAdvisor host check only supports one configured'
' instance.')
self.connection_timeout = int(init_config.get('connection_timeout',
DEFAULT_TIMEOUT))
self.cadvisor_url = None
self.cadvisor_machine_url = None
self.total_mem = 0
self.num_cores = 0
def _parse_machine_info(self, machine_info):
topo_info = machine_info['topology']
# Grab first set of info from return data
topo_info = topo_info[0]
# Store info about total machine memory
if topo_info['memory']:
self.total_mem = topo_info['memory']
self.log.debug("host memory = {}".format(self.total_mem))
else:
self.log.warn("Failed to retrieve host memory size")
# Store information about number of cores (incl. threads)
if machine_info['num_cores']:
self.num_cores = int(machine_info['num_cores'])
self.log.debug("number of cores of machine: {}".format(self.num_cores))
else:
self.log.warn("Failed to retrieve number of cores of host")
def check(self, instance):
if not self.cadvisor_url:
cadvisor_url = instance.get("cadvisor_url", None)
detect_cadvisor_url = instance.get("kubernetes_detect_cadvisor", False)
if not cadvisor_url:
if detect_cadvisor_url:
kubernetes_connector = utils.KubernetesConnector(self.connection_timeout)
host = kubernetes_connector.get_agent_pod_host()
cadvisor_url = "http://{}:4194".format(host)
else:
exception_message = "Either cAdvisor url or kubernetes " \
"detect cAdvisor must be set when " \
"monitoring a Kubernetes Node."
self.log.error(exception_message)
raise Exception(exception_message)
self.cadvisor_url = "{}/{}".format(cadvisor_url, "api/v2.0/stats?count=1")
dimensions = self._set_dimensions(None, instance)
try:
host_metrics = requests.get(self.cadvisor_url, self.connection_timeout).json()
except Exception as e:
self.log.error("Error communicating with cAdvisor to collect data - {}".format(e))
else:
# Retrieve machine info only once
if not self.cadvisor_machine_url:
# Replace path in current cadvisor_url
result = urlparse(self.cadvisor_url)
self.cadvisor_machine_url = urlunparse(result._replace(path="api/v2.0/machine"))
try:
machine_info = requests.get(self.cadvisor_machine_url).json()
except Exception as ex:
self.log.error(
"Error communicating with cAdvisor to collect machine data - {}".format(ex))
else:
self._parse_machine_info(machine_info)
self._parse_send_metrics(host_metrics, dimensions)
def _send_metrics(self, metric_name, value, dimensions, metric_types,
metric_units):
for metric_type in metric_types:
if metric_type == 'rate':
dimensions.update({'unit': metric_units[metric_types.index('rate')]})
self.rate(metric_name + "_sec", value, dimensions)
elif metric_type == 'gauge':
dimensions.update({'unit': metric_units[metric_types.index('gauge')]})
self.gauge(metric_name, value, dimensions)
def _parse_memory(self, memory_data, dimensions):
memory_metrics = METRICS['memory_metrics']
used_mem = -1
for cadvisor_key, (metric_name, metric_types, metric_units) in memory_metrics.items():
if cadvisor_key in memory_data:
self._send_metrics("mem." + metric_name,
memory_data[cadvisor_key],
dimensions,
metric_types, metric_units)
if cadvisor_key == "usage":
used_mem = int(memory_data[cadvisor_key])
# Calculate memory used percent
if used_mem < 0:
self.log.warn("no value for used memory, memory usage (percent) couldn't be calculated")
elif self.total_mem > 0:
used_mem_perc = (float(used_mem) / float(self.total_mem)) * 100
# Send metric percent used
self._send_metrics("mem.used_perc",
used_mem_perc,
dimensions,
metric_types, metric_units)
def _parse_filesystem(self, filesystem_data, dimensions):
filesystem_metrics = METRICS['filesystem_metrics']
for filesystem in filesystem_data:
file_dimensions = dimensions.copy()
file_dimensions['device'] = filesystem['device']
usage_fs = -1
capacity_fs = 0
for cadvisor_key, (metric_name, metric_types,
metric_units) in filesystem_metrics.items():
if cadvisor_key in filesystem:
self._send_metrics("fs." + metric_name,
filesystem[cadvisor_key],
file_dimensions,
metric_types, metric_units)
if cadvisor_key == "usage":
usage_fs = int(filesystem[cadvisor_key])
elif cadvisor_key == "capacity":
capacity_fs = int(filesystem[cadvisor_key])
if usage_fs < 0:
self.log.warn(
"no value for usage size of {}, file system usage (percent) "
"couldn't be calculated".format(
filesystem['device']))
elif capacity_fs > 0:
self._send_metrics("fs.usage_perc",
(float(usage_fs) / capacity_fs) * 100,
file_dimensions,
["gauge"], ["percent"])
else:
self.log.warn(
"no value for capacity of {}, file system usage (percent) "
"couldn't be calculated".format(
filesystem['device']))
def _parse_network(self, network_data, dimensions):
network_interfaces = network_data['interfaces']
network_metrics = METRICS['network_metrics']
interface_sum = {}
# This function is to roll up network metrics for different interfaces
for interface in network_interfaces:
interface_sum = rollup_dictionaries(interface_sum, interface)
network_dimensions = dimensions.copy()
for cadvisor_key, (metric_name, metric_types, metric_units) in network_metrics.items():
if cadvisor_key in interface_sum:
self._send_metrics("net." + metric_name,
interface_sum[cadvisor_key],
network_dimensions,
metric_types,
metric_units)
def _parse_diskio(self, diskio_data, dimensions):
diskio_metrics = METRICS['diskio_metrics']
disk_io_sum = {}
for io_data in diskio_data['io_service_bytes']:
disk_io_sum = rollup_dictionaries(disk_io_sum, io_data['stats'])
for cadvisor_key, (metric_name, metric_types, metric_units) in diskio_metrics.items():
if cadvisor_key in disk_io_sum:
self._send_metrics("io." + metric_name, disk_io_sum[cadvisor_key],
dimensions,
metric_types,
metric_units)
def _parse_cpu(self, cpu_data, dimensions):
cpu_metrics = METRICS['cpu_metrics']
cpu_usage = cpu_data['usage']
for cadvisor_key, (metric_name, metric_types, metric_units) in cpu_metrics.items():
if cadvisor_key in cpu_usage:
# Convert nanoseconds to seconds
cpu_usage_sec = cpu_usage[cadvisor_key] / 1000000000.0
self._send_metrics(
"cpu." + metric_name,
cpu_usage_sec,
dimensions,
metric_types,
metric_units)
# Provide metrics for number of cores if given
if self.num_cores > 0:
self._send_metrics(
"cpu.num_cores",
self.num_cores,
dimensions,
["gauge"],
["number_of_cores"])
def _parse_send_metrics(self, metrics, dimensions):
for host, cadvisor_metrics in metrics.items():
host_dimensions = dimensions.copy()
# Grab first set of metrics from return data
cadvisor_metrics = cadvisor_metrics[0]
if cadvisor_metrics['has_memory'] and cadvisor_metrics['memory']:
self._parse_memory(cadvisor_metrics['memory'], host_dimensions)
if cadvisor_metrics['has_filesystem'] and cadvisor_metrics['filesystem']:
self._parse_filesystem(cadvisor_metrics['filesystem'], host_dimensions)
if cadvisor_metrics['has_network'] and cadvisor_metrics['network']:
self._parse_network(cadvisor_metrics['network'], host_dimensions)
if cadvisor_metrics['has_diskio'] and cadvisor_metrics['diskio']:
self._parse_diskio(cadvisor_metrics['diskio'], host_dimensions)
if cadvisor_metrics['has_cpu'] and cadvisor_metrics['cpu']:
self._parse_cpu(cadvisor_metrics['cpu'], host_dimensions)