# -*- coding: utf-8 -*- # # This file is part of Glances. # # Copyright (C) 2015 Nicolargo # # Glances is free software; you can redistribute it and/or modify # it under the terms of the GNU Lesser General Public License as published by # the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # Glances is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Lesser General Public License for more details. # # You should have received a copy of the GNU Lesser General Public License # along with this program. If not, see . """Docker plugin.""" import numbers import os import re # Import Glances libs from glances.core.glances_timer import getTimeSinceLastUpdate from glances.core.glances_logging import logger from glances.plugins.glances_plugin import GlancesPlugin # Docker-py library (optional and Linux-only) # https://github.com/docker/docker-py try: import docker import requests except ImportError as e: logger.debug("Docker library not found (%s). Glances cannot grab Docker info." % e) docker_tag = False else: docker_tag = True class Plugin(GlancesPlugin): """Glances Docker plugin. stats is a list """ def __init__(self, args=None): """Init the plugin.""" GlancesPlugin.__init__(self, args=args) # The plgin can be disable using: args.disable_docker self.args = args # We want to display the stat in the curse interface self.display_curse = True # Init the Docker API self.docker_client = False def connect(self, version=None): """Connect to the Docker server.""" # Init connection to the Docker API try: if version is None: ret = docker.Client(base_url='unix://var/run/docker.sock') else: ret = docker.Client(base_url='unix://var/run/docker.sock', version=version) except NameError: # docker lib not found return None try: ret.version() except requests.exceptions.ConnectionError as e: # Connexion error (Docker not detected) # Let this message in debug mode logger.debug("Can't connect to the Docker server (%s)" % e) return None except docker.errors.APIError as e: if version is None: # API error (Version mismatch ?) logger.debug("Docker API error (%s)" % e) # Try the connection with the server version version = re.search('server\:\ (.*)\)\".*\)', str(e)) if version: logger.debug("Try connection with Docker API version %s" % version.group(1)) ret = self.connect(version=version.group(1)) else: logger.debug("Can not retreive Docker server version") ret = None else: # API error logger.error("Docker API error (%s)" % e) ret = None except Exception as e: # Others exceptions... # Connexion error (Docker not detected) logger.error("Can't connect to the Docker server (%s)" % e) ret = None # Log an info if Docker plugin is disabled if ret is None: logger.debug("Docker plugin is disable because an error has been detected") return ret def reset(self): """Reset/init the stats.""" self.stats = {} @GlancesPlugin._log_result_decorator def update(self): """Update Docker stats using the input method.""" # Reset stats self.reset() # Get the current Docker API client if not self.docker_client: # First time, try to connect to the server self.docker_client = self.connect() if self.docker_client is None: global docker_tag docker_tag = False # The Docker-py lib is mandatory if not docker_tag or (self.args is not None and self.args.disable_docker): return self.stats if self.input_method == 'local': # Update stats # Exemple: { # "KernelVersion": "3.16.4-tinycore64", # "Arch": "amd64", # "ApiVersion": "1.15", # "Version": "1.3.0", # "GitCommit": "c78088f", # "Os": "linux", # "GoVersion": "go1.3.3" # } self.stats['version'] = self.docker_client.version() # Example: [{u'Status': u'Up 36 seconds', # u'Created': 1420378904, # u'Image': u'nginx:1', # u'Ports': [{u'Type': u'tcp', u'PrivatePort': 443}, # {u'IP': u'0.0.0.0', u'Type': u'tcp', u'PublicPort': 8080, u'PrivatePort': 80}], # u'Command': u"nginx -g 'daemon off;'", # u'Names': [u'/webstack_nginx_1'], # u'Id': u'b0da859e84eb4019cf1d965b15e9323006e510352c402d2f442ea632d61faaa5'}] self.stats['containers'] = self.docker_client.containers() # Get stats for all containers for c in self.stats['containers']: if not hasattr(self, 'docker_stats'): # Create a dict with all the containers' stats instance self.docker_stats = {} # TODO: Find a way to correct this # The following optimization is not compatible with the network stats # The self.docker_client.stats method should be call every time in order to have network stats refreshed # Nevertheless, if we call it every time, Glances is slow... if c['Id'] not in self.docker_stats: # Create the stats instance for the current container try: self.docker_stats[c['Id']] = self.docker_client.stats(c['Id'], decode=True) logger.debug("Create Docker stats object for container {}".format(c['Id'])) except (AttributeError, docker.errors.InvalidVersion) as e: logger.error("Can not call Docker stats method {}".format(e)) # Get the docker stats try: # self.docker_stats[c['Id']] = self.docker_client.stats(c['Id'], decode=True) all_stats = self.docker_stats[c['Id']].next() except: all_stats = {} c['cpu'] = self.get_docker_cpu(c['Id'], all_stats) c['memory'] = self.get_docker_memory(c['Id'], all_stats) # c['network'] = self.get_docker_network(c['Id'], all_stats) elif self.input_method == 'snmp': # Update stats using SNMP # Not available pass return self.stats def get_docker_cpu_old(self, container_id): """Return the container CPU usage by reading /sys/fs/cgroup/. Input: id is the full container id Output: a dict {'total': 1.49, 'user': 0.65, 'system': 0.84} """ ret = {} # Read the stats try: with open('/sys/fs/cgroup/cpuacct/docker/' + container_id + '/cpuacct.stat', 'r') as f: for line in f: m = re.search(r"(system|user)\s+(\d+)", line) if m: ret[m.group(1)] = int(m.group(2)) except IOError as e: logger.error("Can not grab container CPU stat ({0})".format(e)) return ret if isinstance(ret["system"], numbers.Number) and isinstance(ret["user"], numbers.Number): ret["total"] = ret["system"] + ret["user"] # Return the stats return ret def get_docker_cpu(self, container_id, all_stats): """Return the container CPU usage. Input: id is the full container id all_stats is the output of the stats method of the Docker API Output: a dict {'total': 1.49} """ cpu_new = {} ret = {'total': 0.0} # Read the stats # For each container, you will find a pseudo-file cpuacct.stat, # containing the CPU usage accumulated by the processes of the container. # Those times are expressed in ticks of 1/USER_HZ of a second. # On x86 systems, USER_HZ is 100. try: cpu_new['total'] = all_stats['cpu_stats']['cpu_usage']['total_usage'] cpu_new['system'] = all_stats['cpu_stats']['system_cpu_usage'] cpu_new['nb_core'] = len(all_stats['cpu_stats']['cpu_usage']['percpu_usage']) except KeyError as e: # all_stats do not have CPU information logger.debug("Can not grab CPU usage for container {0} ({1}). Trying fallback method.".format(container_id, e)) # Trying fallback to old grab method ret = self.get_docker_cpu_old(container_id) # Get the user ticks ticks = self.get_user_ticks() for k in ret.keys(): ret[k] = float(ret[k]) / ticks else: # Previous CPU stats stored in the cpu_old variable if not hasattr(self, 'cpu_old'): # First call, we init the cpu_old variable self.cpu_old = {} try: self.cpu_old[container_id] = cpu_new except (IOError, UnboundLocalError): pass if container_id not in self.cpu_old: try: self.cpu_old[container_id] = cpu_new except (IOError, UnboundLocalError): pass else: # cpu_delta = float(cpu_new['total'] - self.cpu_old[container_id]['total']) system_delta = float(cpu_new['system'] - self.cpu_old[container_id]['system']) if cpu_delta > 0.0 and system_delta > 0.0: ret['total'] = (cpu_delta / system_delta) * float(cpu_new['nb_core']) * 100 # Save stats to compute next stats self.cpu_old[container_id] = cpu_new # Return the stats return ret def get_docker_memory_old(self, container_id): """Return the container MEMORY usage by reading /sys/fs/cgroup/. Input: id is the full container id Output: a dict {'rss': 1015808, 'cache': 356352} """ ret = {} # Read the stats try: with open('/sys/fs/cgroup/memory/docker/' + container_id + '/memory.stat', 'r') as f: for line in f: m = re.search(r"(rss|cache)\s+(\d+)", line) if m: ret[m.group(1)] = int(m.group(2)) except IOError as e: logger.error("Can not grab container MEM stat ({0})".format(e)) return ret # Return the stats return ret def get_docker_memory(self, container_id, all_stats): """Return the container MEMORY. Input: id is the full container id all_stats is the output of the stats method of the Docker API Output: a dict {'rss': 1015808, 'cache': 356352, 'usage': ..., 'max_usage': ...} """ ret = {} # Read the stats try: ret['rss'] = all_stats['memory_stats']['stats']['rss'] ret['cache'] = all_stats['memory_stats']['stats']['cache'] ret['usage'] = all_stats['memory_stats']['usage'] ret['max_usage'] = all_stats['memory_stats']['max_usage'] except KeyError as e: # all_stats do not have MEM information logger.debug("Can not grab MEM usage for container {0} ({1}). Trying fallback method.".format(container_id, e)) # Trying fallback to old grab method ret = self.get_docker_memory_old(container_id) # Return the stats return ret def get_docker_network(self, container_id, all_stats): """Return the container network usage using the Docker API (v1.0 or higher). Input: id is the full container id Output: a dict {'time_since_update': 3000, 'rx': 10, 'tx': 65}. """ # Init the returned dict network_new = {} # Read the rx/tx stats (in bytes) try: netiocounters = all_stats["network"] except KeyError as e: # all_stats do not have NETWORK information logger.debug("Can not grab NET usage for container {0} ({1})".format(container_id, e)) # No fallback available... return network_new # Previous network interface stats are stored in the network_old variable if not hasattr(self, 'netiocounters_old'): # First call, we init the network_old var self.netiocounters_old = {} try: self.netiocounters_old[container_id] = netiocounters except (IOError, UnboundLocalError): pass if container_id not in self.netiocounters_old: try: self.netiocounters_old[container_id] = netiocounters except (IOError, UnboundLocalError): pass else: # By storing time data we enable Rx/s and Tx/s calculations in the # XML/RPC API, which would otherwise be overly difficult work # for users of the API network_new['time_since_update'] = getTimeSinceLastUpdate('docker_net_{}'.format(container_id)) network_new['rx'] = netiocounters["rx_bytes"] - self.netiocounters_old[container_id]["rx_bytes"] network_new['tx'] = netiocounters["tx_bytes"] - self.netiocounters_old[container_id]["tx_bytes"] network_new['cumulative_rx'] = netiocounters["rx_bytes"] network_new['cumulative_tx'] = netiocounters["tx_bytes"] # Save stats to compute next bitrate self.netiocounters_old[container_id] = netiocounters # Return the stats return network_new def get_user_ticks(self): """Return the user ticks by reading the environment variable.""" return os.sysconf(os.sysconf_names['SC_CLK_TCK']) def msg_curse(self, args=None): """Return the dict to display in the curse interface.""" # Init the return message ret = [] # Only process if stats exist (and non null) and display plugin enable... if not self.stats or args.disable_docker or len(self.stats['containers']) == 0: return ret # Build the string message # Title msg = '{0}'.format('CONTAINERS') ret.append(self.curse_add_line(msg, "TITLE")) msg = ' {0}'.format(len(self.stats['containers'])) ret.append(self.curse_add_line(msg)) msg = ' ({0} {1})'.format('served by Docker', self.stats['version']["Version"]) ret.append(self.curse_add_line(msg)) ret.append(self.curse_new_line()) # Header ret.append(self.curse_new_line()) msg = '{0:>14}'.format('Id') ret.append(self.curse_add_line(msg)) msg = ' {0:20}'.format('Name') ret.append(self.curse_add_line(msg)) msg = '{0:>26}'.format('Status') ret.append(self.curse_add_line(msg)) msg = '{0:>6}'.format('CPU%') ret.append(self.curse_add_line(msg)) msg = '{0:>7}'.format('MEM') ret.append(self.curse_add_line(msg)) # msg = '{0:>6}'.format('Rx/s') # ret.append(self.curse_add_line(msg)) # msg = '{0:>6}'.format('Tx/s') # ret.append(self.curse_add_line(msg)) msg = ' {0:8}'.format('Command') ret.append(self.curse_add_line(msg)) # Data for container in self.stats['containers']: ret.append(self.curse_new_line()) # Id msg = '{0:>14}'.format(container['Id'][0:12]) ret.append(self.curse_add_line(msg)) # Name name = container['Names'][0] if len(name) > 20: name = '_' + name[:-19] else: name = name[:20] msg = ' {0:20}'.format(name) ret.append(self.curse_add_line(msg)) # Status status = self.container_alert(container['Status']) msg = container['Status'].replace("minute", "min") msg = '{0:>26}'.format(msg[0:25]) ret.append(self.curse_add_line(msg, status)) # CPU try: msg = '{0:>6.1f}'.format(container['cpu']['total']) except KeyError: msg = '{0:>6}'.format('?') ret.append(self.curse_add_line(msg)) # MEM try: msg = '{0:>7}'.format(self.auto_unit(container['memory']['usage'])) except KeyError: msg = '{0:>7}'.format('?') ret.append(self.curse_add_line(msg)) # NET RX/TX # for r in ['rx', 'tx']: # try: # value = self.auto_unit(int(container['network'][r] // container['network']['time_since_update'] * 8)) + "b" # msg = '{0:>6}'.format(value) # except KeyError: # msg = '{0:>6}'.format('?') # ret.append(self.curse_add_line(msg)) # Command msg = ' {0}'.format(container['Command']) ret.append(self.curse_add_line(msg)) return ret def container_alert(self, status): """Analyse the container status.""" if "Paused" in status: return 'CAREFUL' else: return 'OK'