def check_backlogged_devices(self):
""""Checks the status of backlogged devices.
Skips newly spun up instances during their booting time as specified
in the boot time parameter.
:return A dict of the format:
{'reachable': [<hd_id>,..], 'dead': [<hd_id>,..]}
"""
response_dict = {'reachable': [], 'dead': []}
LOG.debug("Current Backlogged devices: %s",
self.backlog_devices.keys())
for hd_id in self.backlog_devices.keys():
hd = self.backlog_devices[hd_id]
if hd.get('mgmt_url'):
if not timeutils.is_older_than(hd['created_at'],
BOOT_TIME_INTERVAL):
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s hasn't "
"passed minimum boot time. Skipping it. ",
{'hd_id': hd_id, 'ip': hd['mgmt_url'].get('ip_address', None)})
continue
LOG.debug("Checking device: %(hd_id)s @ %(ip)s for "
"reachability.", {'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)})
if _is_pingable(hd['mgmt_url'].get('ip_address', None)):
response_dict['reachable'].append(hd_id)
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s is now "
"reachable. Adding it to response",
{'hd_id': hd_id, 'ip': hd['mgmt_url'].get('ip_address', None)})
else:
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s still not "
"reachable ", {'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)})
if hd.get('backlog_insertion_ts'):
if timeutils.is_older_than(
hd['backlog_insertion_ts'],
cfg.CONF.servicevm_agent.device_dead_timeout):
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s hasn't "
"been reachable for the last %(time)d seconds. "
"Marking it dead.",
{'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None),
'time': cfg.CONF.servicevm_agent.
device_dead_timeout})
response_dict['dead'].append(hd_id)
else:
response_dict['dead'].append(hd_id)
else:
response_dict['dead'].append(hd_id)
return response_dict
def check_backlogged_hosting_devices(self):
""""Checks the status of backlogged hosting devices.
Skips newly spun up instances during their booting time as specified
in the boot time parameter.
:return A dict of the format:
{'reachable': [<hd_id>,..], 'dead': [<hd_id>,..]}
"""
response_dict = {'reachable': [], 'dead': []}
LOG.debug("Current Backlogged hosting devices: %s",
self.backlog_hosting_devices.keys())
for hd_id in self.backlog_hosting_devices.keys():
hd = self.backlog_hosting_devices[hd_id]['hd']
if not timeutils.is_older_than(hd['created_at'],
hd['booting_time']):
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s hasn't passed "
"minimum boot time. Skipping it. "),
{'hd_id': hd_id, 'ip': hd['management_ip_address']})
continue
LOG.info(_("Checking hosting device: %(hd_id)s @ %(ip)s for "
"reachability."), {'hd_id': hd_id,
'ip': hd['management_ip_address']})
if _is_pingable(hd['management_ip_address']):
hd.pop('backlog_insertion_ts', None)
del self.backlog_hosting_devices[hd_id]
response_dict['reachable'].append(hd_id)
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s is now "
"reachable. Adding it to response"),
{'hd_id': hd_id, 'ip': hd['management_ip_address']})
else:
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s still not "
"reachable "), {'hd_id': hd_id,
'ip': hd['management_ip_address']})
if timeutils.is_older_than(
hd['backlog_insertion_ts'],
cfg.CONF.cfg_agent.hosting_device_dead_timeout):
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s hasn't "
"been reachable for the last %(time)d seconds. "
"Marking it dead.",
{'hd_id': hd_id,
'ip': hd['management_ip_address'],
'time': cfg.CONF.cfg_agent.
hosting_device_dead_timeout})
response_dict['dead'].append(hd_id)
hd.pop('backlog_insertion_ts', None)
del self.backlog_hosting_devices[hd_id]
LOG.debug("Response: %s", response_dict)
return response_dict
def check_backlogged_hosting_devices(self):
""""Checks the status of backlogged hosting devices.
Skips newly spun up instances during their booting time as specified
in the boot time parameter.
:return A dict of the format:
{'reachable': [<hd_id>,..], 'dead': [<hd_id>,..]}
"""
response_dict = {'reachable': [], 'dead': []}
LOG.debug("Current Backlogged hosting devices: %s",
self.backlog_hosting_devices.keys())
for hd_id in self.backlog_hosting_devices.keys():
hd = self.backlog_hosting_devices[hd_id]['hd']
if not timeutils.is_older_than(hd['created_at'],
hd['booting_time']):
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s hasn't passed "
"minimum boot time. Skipping it. "),
{'hd_id': hd_id, 'ip': hd['management_ip_address']})
continue
LOG.info(_("Checking hosting device: %(hd_id)s @ %(ip)s for "
"reachability."), {'hd_id': hd_id,
'ip': hd['management_ip_address']})
if _is_pingable(hd['management_ip_address']):
hd.pop('backlog_insertion_ts', None)
del self.backlog_hosting_devices[hd_id]
response_dict['reachable'].append(hd_id)
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s is now "
"reachable. Adding it to response"),
{'hd_id': hd_id, 'ip': hd['management_ip_address']})
else:
LOG.info(_("Hosting device: %(hd_id)s @ %(ip)s still not "
"reachable "), {'hd_id': hd_id,
'ip': hd['management_ip_address']})
if timeutils.is_older_than(
hd['backlog_insertion_ts'],
cfg.CONF.hosting_device_dead_timeout):
LOG.debug("Hosting device: %(hd_id)s @ %(ip)s hasn't "
"been reachable for the last %(time)d seconds. "
"Marking it dead.",
{'hd_id': hd_id,
'ip': hd['management_ip_address'],
'time': cfg.CONF.hosting_device_dead_timeout})
response_dict['dead'].append(hd_id)
hd.pop('backlog_insertion_ts', None)
del self.backlog_hosting_devices[hd_id]
LOG.debug("Response: %s", response_dict)
return response_dict
def _process_static_route(self, ip, info):
heart_beat_time = info['time']
flag = info['flag']
static_route = {
"next_hop_address": ip,
"priority":"60",
"description":"SLB static route"
}
if timeutils.is_older_than(heart_beat_time,
self.conf.haproxy_int.proxy_down_time):
for external_network in self.external_networks:
strs = external_network.split('/')
static_route["ip_address"] = strs[0]
static_route["mask_length"] = strs[1]
success = self.ngfw_client.clear_static_route(static_route)
if not success:
LOG.warn(_('Failed to remove static route to proxy %s'), ip)
return
del self.proxy_ips[ip]
LOG.debug(_('Remove static route to proxy %s'), ip)
else:
if not flag:
for external_network in self.external_networks:
strs = external_network.split('/')
static_route["ip_address"] = strs[0]
static_route["mask_length"] = strs[1]
success = self.ngfw_client.make_static_route(static_route)
if not success:
LOG.warn(_('Failed to add static route to proxy %s'), ip)
return
info['flag'] = True
LOG.debug(_('Add static route to proxy %s'), ip)
def monitor_agent_state(self):
"""Represents a thread which maintains list of active
and inactive agents based on the heartbeat recorded
"""
#Do nothing until plugin is initialized
agents_to_process = []
if not self.plugin:
status = self.get_plugin_and_initialize()
if not status:
LOG.debug("Plugin not defined...returning")
return
if not self.agent_ext_support:
LOG.debug("Agent extension is not loaded by plugin")
return
try:
self.agents = self.plugin.get_agents(self.context,
filters={'agent_type': [AGENT_TYPE_OVSVAPP]})
except Exception:
LOG.exception(_("Unable to get agent list continue..."))
return
for agent in self.agents:
agent_time_stamp = agent['heartbeat_timestamp']
agent_id = agent['id']
status = timeutils.is_older_than(agent_time_stamp,
cfg.CONF.agent_down_time * 2)
LOG.debug(_("for agent %(agent)s agent_state %(state)s"),
{'agent': agent, 'state': status})
try:
if not status:
if agent_id not in self.active_agents:
LOG.debug("Moving agent: %s from inactive to "
"active", agent_id)
self.active_agents.append(agent_id)
self._update_agent_state(agent_id, True)
if agent_id in self.inactive_agents:
LOG.debug("Removing agent: %s from inactive "
"agent list", agent_id)
self.inactive_agents.remove(agent_id)
else:
if agent_id not in self.inactive_agents:
LOG.debug("Moving agent: %s from active to "
"inactive", agent_id)
self.inactive_agents.append(agent_id)
if self._update_agent_state(agent_id, False):
agents_to_process.append(agent)
if agent_id in self.active_agents:
LOG.debug("Removing agent: %s from active "
"agent list", agent_id)
self.active_agents.remove(agent_id)
except Exception:
LOG.exception(_("Exception occurred in monitor_agent_state.."))
LOG.debug("Number of agents for processing: %s",
len(agents_to_process))
for agent in agents_to_process:
self.process_ovsvapp_agent(agent)
return
def _process_static_route(self, ip, info):
heart_beat_time = info['time']
flag = info['flag']
static_route = {
"next_hop_address": ip,
"priority": "60",
"description": "SLB static route"
}
if timeutils.is_older_than(heart_beat_time,
self.conf.haproxy_int.proxy_down_time):
for external_network in self.external_networks:
strs = external_network.split('/')
static_route["ip_address"] = strs[0]
static_route["mask_length"] = strs[1]
success = self.ngfw_client.clear_static_route(static_route)
if not success:
LOG.warn(_('Failed to remove static route to proxy %s'),
ip)
return
del self.proxy_ips[ip]
LOG.debug(_('Remove static route to proxy %s'), ip)
else:
if not flag:
for external_network in self.external_networks:
strs = external_network.split('/')
static_route["ip_address"] = strs[0]
static_route["mask_length"] = strs[1]
success = self.ngfw_client.make_static_route(static_route)
if not success:
LOG.warn(_('Failed to add static route to proxy %s'),
ip)
return
info['flag'] = True
LOG.debug(_('Add static route to proxy %s'), ip)
def is_agent_down(cls, heart_beat_time):
return timeutils.is_older_than(heart_beat_time,
cfg.CONF.agent_down_time)
def is_agent_down(cls, heart_beat_time):
return timeutils.is_older_than(heart_beat_time,
cfg.CONF.agent_down_time)
def is_agent_down(cls, heart_beat_time,
timeout=cfg.CONF.general.cfg_agent_down_time):
return timeutils.is_older_than(heart_beat_time, timeout)
def check_backlogged_devices(self):
""""Checks the status of backlogged devices.
Skips newly spun up instances during their booting time as specified
in the boot time parameter.
:return A dict of the format:
{'reachable': [<hd_id>,..], 'dead': [<hd_id>,..]}
"""
response_dict = {'reachable': [], 'dead': []}
LOG.debug("Current Backlogged devices: %s",
self.backlog_devices.keys())
for hd_id in self.backlog_devices.keys():
hd = self.backlog_devices[hd_id]
if hd.get('mgmt_url'):
if not timeutils.is_older_than(hd['created_at'],
BOOT_TIME_INTERVAL):
LOG.debug(
"Hosting device: %(hd_id)s @ %(ip)s hasn't "
"passed minimum boot time. Skipping it. ", {
'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)
})
continue
LOG.debug(
"Checking device: %(hd_id)s @ %(ip)s for "
"reachability.", {
'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)
})
if _is_pingable(hd['mgmt_url'].get('ip_address', None)):
response_dict['reachable'].append(hd_id)
LOG.debug(
"Hosting device: %(hd_id)s @ %(ip)s is now "
"reachable. Adding it to response", {
'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)
})
else:
LOG.debug(
"Hosting device: %(hd_id)s @ %(ip)s still not "
"reachable ", {
'hd_id': hd_id,
'ip': hd['mgmt_url'].get('ip_address', None)
})
if hd.get('backlog_insertion_ts'):
if timeutils.is_older_than(
hd['backlog_insertion_ts'],
cfg.CONF.servicevm_agent.device_dead_timeout):
LOG.debug(
"Hosting device: %(hd_id)s @ %(ip)s hasn't "
"been reachable for the last %(time)d seconds. "
"Marking it dead.", {
'hd_id':
hd_id,
'ip':
hd['mgmt_url'].get('ip_address', None),
'time':
cfg.CONF.servicevm_agent.
device_dead_timeout
})
response_dict['dead'].append(hd_id)
else:
response_dict['dead'].append(hd_id)
else:
response_dict['dead'].append(hd_id)
return response_dict
def is_agent_down(cls, heart_beat_time,
timeout=cfg.CONF.general.cfg_agent_down_time):
return timeutils.is_older_than(heart_beat_time, timeout)