def unregi_url(url):
try:
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL + ' WHERE url = \'' + url + '\''
with DB.connection() as conn:
url_info = conn.cursor().execute(sql).fetchall()
conn.close()
# if no exist
if len(url_info) == 0:
res_body = {'Result': 'SUCCESS'}
else:
# delete db
sql = 'DELETE FROM ' + DB.REGI_SYS_TBL + ' WHERE url = \'' + url + '\''
ret = DB.sql_execute(sql)
if ret == 'SUCCESS':
res_body = {'Result': 'SUCCESS'}
else:
res_body = {'Result': 'FAIL'}
return res_body
except:
LOG.exception()
return {'Result': 'FAIL'}
def proc_dis_system(node, dummy):
try:
result = dict()
for sys_type in CONF.watchdog()['check_system']:
event_list = DB.get_event_list(sys_type)
sql = 'SELECT ' + DB.STATUS_TBL + '.nodename, ' + DB.NODE_INFO_TBL + '.ip_addr, ' + ", ".join(event_list) + ' FROM ' + DB.STATUS_TBL + \
' INNER JOIN ' + DB.NODE_INFO_TBL + ' ON ' + DB.STATUS_TBL + '.nodename = ' + DB.NODE_INFO_TBL + '.nodename WHERE type = \'' + sys_type + '\''
if not node == 'all':
sql = sql + ' and ' + DB.NODE_INFO_TBL + '.nodename = \'' + node + '\''
with DB.connection() as conn:
nodes_info = conn.cursor().execute(sql).fetchall()
conn.close()
for row in nodes_info:
line = dict()
line['TYPE'] = sys_type
line['IP'] = row[1]
i = 2
for item in event_list:
line[item] = row[i]
i = i + 1
result[row[0]] = line
return result
except:
LOG.exception()
return {'Result': 'FAIL'}
def regi_url(url, auth):
try:
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL + ' WHERE url = \'' + url + '\''
with DB.connection() as conn:
url_info = conn.cursor().execute(sql).fetchall()
conn.close()
# if already exist
if len(url_info) == 1:
res_body = {'Result': 'SUCCESS'}
else:
# insert db
sql = 'INSERT INTO ' + DB.REGI_SYS_TBL + ' VALUES (\'' + url + '\', \'' + auth + '\' )'
ret = DB.sql_execute(sql)
if ret == 'SUCCESS':
res_body = {'Result': 'SUCCESS'}
else:
res_body = {'Result': 'FAIL'}
return res_body
except:
LOG.exception()
return {'Result': 'FAIL'}
def get_event_list(url, auth):
try:
sql_evt = 'SELECT * FROM ' + DB.EVENT_TBL
with DB.connection() as conn:
evt_list = conn.cursor().execute(sql_evt).fetchall()
conn.close()
event_list = []
for nodename, item, grade, pre_grade, reason, time in evt_list:
evt = {
'event': 'occur',
'system': nodename,
'item': item,
'grade': grade,
'pre_grade': pre_grade,
'reason': 'fail_reason',
'time': time
}
event_list.append(evt)
res_body = {'Result': 'SUCCESS', 'Event list': event_list}
return res_body
except:
LOG.exception()
return {'Result': 'FAIL'}
def push_event(node_name, item, grade, desc, time):
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL
with DB.connection() as conn:
url_list = conn.cursor().execute(sql).fetchall()
conn.close()
for url, auth in url_list:
header = {'Content-Type': 'application/json', 'Authorization': auth}
req_body = {
'event': 'occur',
'system': node_name,
'item': item,
'grade': grade,
'desc': desc,
'time': time
}
req_body_json = json.dumps(req_body)
try:
requests.post(url, headers=header, data=req_body_json, timeout=2)
except:
# rest timeout
LOG.exception()
def xos_status_check(conn, db_log, node_name):
xos_status = 'ok'
xos_list = []
fail_reason = []
try:
url = CONF.xos()['xos_rest_server']
account = CONF.xos()['xos_rest_account']
cmd = 'curl -H "Accept: application/json; indent=4" -u ' + account + ' -X GET ' + url + '/api/core/xoses/'
result = Popen(cmd, stdout=PIPE, stderr=PIPE, shell=True)
output, error = result.communicate()
if result.returncode != 0:
LOG.error("Cmd Fail, cause => %s", error)
return 'fail', None
xos_array = json.loads(output)
for xos_info in xos_array:
backend_status = xos_info['backend_status']
LOG.info('xos_status_backend_status = ' + backend_status)
tmp = str(backend_status).split('-')
if tmp[0].strip() == '0':
status = 'ok'
else:
status = 'nok'
xos_json = {
'name': xos_info['name'],
'status': status,
'description': tmp[1].strip()
}
xos_list.append(xos_json)
if status == 'nok':
xos_status = 'nok'
fail_reason.append(xos_json)
try:
sql = 'UPDATE ' + DB.XOS_TBL + \
' SET xos_status = \"' + str(xos_list) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE XOS STATUS INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] XOS STATUS DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
xos_status = 'fail'
return xos_status, fail_reason
def regi_url(url, auth):
try:
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL + ' WHERE url = \'' + url + '\''
sql_evt = 'SELECT * FROM ' + DB.EVENT_TBL
with DB.connection() as conn:
url_info = conn.cursor().execute(sql).fetchall()
evt_list = conn.cursor().execute(sql_evt).fetchall()
conn.close()
event_list = []
for nodename, item, grade, desc, time in evt_list:
if not grade in ['ok', 'normal']:
evt = {
'event': 'occur',
'system': nodename,
'item': item,
'grade': grade,
'desc': desc,
'time': time
}
event_list.append(evt)
# if already exist
if len(url_info) == 1:
res_body = {'Result': 'SUCCESS', 'Event list': event_list}
else:
# insert db
sql = 'INSERT INTO ' + DB.REGI_SYS_TBL + ' VALUES (\'' + url + '\', \'' + auth + '\' )'
ret = DB.sql_execute(sql)
if ret == 'SUCCESS':
res_body = {'Result': 'SUCCESS', 'Event list': event_list}
else:
res_body = {'Result': 'FAIL'}
return res_body
except:
LOG.exception()
return {'Result': 'FAIL'}
def onos_ha_check(conn, db_log):
try:
stats_url = CONF.ha()['ha_proxy_server']
account = CONF.ha()['ha_proxy_account']
cmd = 'curl --user ' + account + ' --header \'Accept: text/html, application/xhtml+xml, image/jxr, */*\' \"' + stats_url + '\"'
result = Popen(cmd, stdout=PIPE, stderr=PIPE, shell=True)
output, error = result.communicate()
if result.returncode != 0:
LOG.error("Cmd Fail, cause => %s", error)
return None
else:
report_data = csv.DictReader(output.lstrip('# ').splitlines())
dic_stat = dict()
for row in report_data:
if row['pxname'].strip() == 'stats' or row['svname'].strip(
) == 'BACKEND':
continue
dtl_list = {
'name': row['svname'],
'req_count': row['stot'],
'succ_count': row['hrsp_2xx'],
'node_sts': row['status']
}
svc_type = row['pxname']
if (dic_stat.has_key(svc_type)):
dic_stat[svc_type].append(dtl_list)
else:
dic_stat[svc_type] = list()
dic_stat[svc_type].append(dtl_list)
try:
str_dic_stat = str(dic_stat)
sql = 'UPDATE ' + DB.HA_TBL + \
' SET stats = \"' + str_dic_stat + '\"' + \
' WHERE ha_key = \"' + 'HA' + '\"'
db_log.write_log('----- UPDATE HA INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] HA DB Update Fail.')
except:
LOG.exception()
return dic_stat
except:
LOG.exception()
return None
def push_event(node_name, item, grade, pre_grade, reason, time, flush_alarm):
global history_log
try:
history_log.write_log('[%s][%s][%s->%s] %s', node_name, item,
pre_grade, grade, reason)
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL
with DB.connection() as conn:
url_list = conn.cursor().execute(sql).fetchall()
conn.close()
for url, auth in url_list:
header = {
'Content-Type': 'application/json',
'Authorization': str(auth)
}
req_body = {
'system': node_name,
'item': item,
'grade': grade,
'pre_grade': pre_grade,
'reason': reason,
'time': time
}
req_body_json = json.dumps(req_body)
try:
requests.post(str(url),
headers=header,
data=req_body_json,
timeout=2)
except:
# Push event does not respond
pass
reason_str = ''
if type(reason) == list:
if len(reason) > 0:
reason_str = '-- ' + '\n-- '.join(reason)
else:
reason_str = str(reason)
ALARM.queue_alarm(node_name + ' ' + item + ' ' + grade.upper(),
reason_str, time)
if flush_alarm:
ALARM.flush_pending_alarm()
except:
LOG.exception()
def occur_event(conn, node_name, item, pre_value, cur_value):
time = str(datetime.now())
desc = pre_value + ' -> ' + cur_value
sql = 'UPDATE ' + DB.EVENT_TBL + \
' SET grade = \'' + cur_value + '\'' + ',' + \
' desc = \'' + desc + '\'' + ',' + \
' time = \'' + time + '\'' + \
' WHERE nodename = \'' + node_name + '\' and item = \'' + item + '\''
LOG.info('Update alarm info = ' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
LOG.error('DB Update Fail.')
push_event(node_name, item, cur_value, desc, time)
def get_node_list(nodes, param, tbl=DB.NODE_INFO_TBL):
try:
if nodes == 'all':
sql = 'SELECT ' + param + ' FROM ' + tbl
else:
sql = 'SELECT ' + param + ' FROM ' + tbl + ' WHERE nodename = \'' + nodes + '\''
with DB.connection() as conn:
nodes_info = conn.cursor().execute(sql).fetchall()
conn.close()
return nodes_info
except:
LOG.exception()
return None
def proc_dis_ha(dummy, param):
try:
sql = 'SELECT stats FROM ' + DB.HA_TBL + ' WHERE ha_key = \'HA\''
with DB.connection() as conn:
nodes_info = conn.cursor().execute(sql).fetchone()
conn.close()
for value in nodes_info:
return json.loads(str(value).replace('\'', '\"'))
return {'HA': 'FAIL'}
except:
LOG.exception()
return {'Result': 'FAIL'}
def run(self):
# DB initiation
DB.db_initiation()
# Start RESTful server
try:
REST_SVR.rest_server_start()
except:
print 'Rest Server failed to start'
LOG.exception()
sys.exit(1)
# Periodic monitoring
if CONF.watchdog()['interval'] == 0:
LOG.info("--- Not running periodic monitoring ---")
while True:
time.sleep(3600)
else:
LOG.info("--- Periodic Monitoring Start ---")
conn = DB.connection()
while True:
try:
watchdog.periodic(conn)
time.sleep(CONF.watchdog()['interval'])
except:
watchdog.push_event('sonawatcher', 'disconnect',
'critical',
'sonawatcher server shutdown',
str(datetime.now()))
conn.close()
LOG.exception()
sys.exit(1)
def occur_event(conn, db_log, node_name, item, pre_grade, cur_grade, reason):
try:
time = str(datetime.now())
sql = 'UPDATE ' + DB.EVENT_TBL + \
' SET grade = \'' + cur_grade + '\'' + ',' + \
' pre_grade = \'' + pre_grade + '\'' + ',' + \
' reason = \"' + str(reason) + '\"' + ',' + \
' time = \'' + time + '\'' + \
' WHERE nodename = \'' + node_name + '\' and item = \'' + item + '\''
db_log.write_log('----- UPDATE EVENT INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] EVENT INFO DB Update Fail.')
push_event(node_name, item, cur_grade, pre_grade, reason, time, False)
except:
LOG.exception()
def push_event(node_name, item, grade, pre_grade, reason, time):
global history_log
try:
history_log.write_log('[%s][%s][%s][%s] %s', node_name, item, grade,
pre_grade, reason)
sql = 'SELECT * FROM ' + DB.REGI_SYS_TBL
with DB.connection() as conn:
url_list = conn.cursor().execute(sql).fetchall()
conn.close()
for url, auth in url_list:
header = {
'Content-Type': 'application/json',
'Authorization': str(auth)
}
req_body = {
'system': node_name,
'item': item,
'grade': grade,
'pre_grade': pre_grade,
'reason': reason,
'time': time
}
req_body_json = json.dumps(req_body)
try:
requests.post(str(url),
headers=header,
data=req_body_json,
timeout=2)
except:
# Push event does not respond
pass
except:
LOG.exception()
def check_resource(conn, db_log, node_name, user_name, node_ip):
try:
cpu = str(get_cpu_usage(user_name, node_ip, True))
mem = str(get_mem_usage(user_name, node_ip, True))
disk = str(get_disk_usage(user_name, node_ip, True))
try:
sql = 'UPDATE ' + DB.RESOURCE_TBL + \
' SET cpu = \'' + cpu + '\',' + \
' memory = \'' + mem + '\',' + \
' disk = \'' + disk + '\'' \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE RESOURCE INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] RESOURCE DB Update Fail.')
except:
LOG.exception()
return cpu, mem, disk
except:
LOG.exception()
return -1, -1, -1
def swarm_check(conn, db_log, node_name, user_name, node_ip):
str_node = ''
str_service = ''
str_ps = ''
ret_app = 'ok'
ret_node = 'ok'
node_rt = SshCommand.ssh_exec(user_name, node_ip, 'sudo docker node ls')
if node_rt is not None:
try:
leader_flag = False
for line in node_rt.splitlines():
line = line.decode('utf-8')
str_node = str_node + line + '\n'
if line.startswith('ID'):
continue
if 'Leader' in line:
leader_flag = True
if not ('Ready' in line and 'Active' in line):
ret_node = 'nok'
break
if 'Down' in line:
ret_node = 'nok'
break
if not leader_flag:
ret_node = 'nok'
except:
LOG.exception()
ret_node = 'nok'
else:
LOG.error("\'%s\' Swarm Node Check Error", node_ip)
str_node = 'fail'
service_rt = SshCommand.ssh_exec(user_name, node_ip,
'sudo docker service ls')
if service_rt is not None:
try:
for app in CONF.swarm()['app_list']:
find_flag = False
for line in service_rt.splitlines():
line = line.decode('utf-8')
if line.startswith('ID'):
continue
id, name, mode, rep, img = line.split()
if app == name:
find_flag = True
rep_tmp = rep.split('/')
if not (rep_tmp[0] == rep_tmp[1]):
ret_app = 'nok'
break
if not find_flag:
ret_app = 'nok'
break
except:
LOG.exception()
ret_app = 'nok'
for line in service_rt.splitlines():
line = line.decode('utf-8')
str_service = str_service + line + '\n'
else:
LOG.error("\'%s\' Swarm Service Check Error", node_ip)
str_service = 'fail'
ret_app = 'nok'
try:
for app in CONF.swarm()['app_list']:
ps_rt = SshCommand.ssh_exec(user_name, node_ip,
'sudo docker service ps ' + app)
str_ps = str_ps + ' * ' + app + '\n\n'
if ps_rt is not None:
for line in ps_rt.splitlines():
line = line.decode('utf-8')
str_ps = str_ps + line + '\n'
else:
LOG.error("\'%s\' Swarm PS Check Error", node_ip)
str_ps = str_ps + 'Command failure(' + app + ')\n'
str_ps = str_ps + '\n'
except:
LOG.exception()
try:
sql = 'UPDATE ' + DB.SWARM_TBL + \
' SET node = \'' + str_node + '\',' + \
' service = \'' + str_service + '\',' + \
' ps = \'' + str_ps + '\'' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE SWARM INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] SWARN DB Update Fail.')
except:
LOG.exception()
return ret_app, ret_node
def swarm_node_check(conn, db_log, node_name, username, node_ip,
swarm_manager):
node_status = 'ok'
node_list = []
fail_reason = []
try:
cmd = 'ssh root@' + swarm_manager + ' \"sudo docker node ls\"'
node_rt = SshCommand.ssh_exec(username, node_ip, cmd)
if node_rt is not None:
try:
leader_flag = False
for line in node_rt.splitlines():
line = line.decode('utf-8')
line = " ".join(line.replace('*', '').split())
tmp = line.split(' ')
if line.startswith('ID'):
continue
if 'Leader' in line:
node_json = {
'hostname': tmp[1],
'status': tmp[2],
'availability': tmp[3],
'manager': tmp[4]
}
leader_flag = True
if not ('Ready' in line and 'Active' in line):
node_status = 'nok'
fail_reason.append(tmp[1] + ' node is not ready.')
else:
node_json = {
'hostname': tmp[1],
'status': tmp[2],
'availability': tmp[3],
'manager': ''
}
if 'Down' in line:
node_status = 'nok'
fail_reason.append(tmp[1] + ' node is down.')
node_list.append(node_json)
if not leader_flag:
node_status = 'nok'
fail_reason.append('swarm leader node does not exist.')
except:
LOG.exception()
node_status = 'nok'
else:
LOG.error("\'%s\' Swarm Node Check Error", node_ip)
node_status = 'fail'
try:
sql = 'UPDATE ' + DB.SWARM_TBL + \
' SET node = \"' + str(node_list) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE SWARM NODE INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] SWARM NODE DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
node_status = 'fail'
return node_status, fail_reason
def swarm_service_check(conn, db_log, node_name, username, node_ip,
swarm_manager):
service_status = 'ok'
service_list = []
ps_list = []
fail_reason = []
try:
cmd = 'ssh root@' + swarm_manager + ' \"sudo docker service ls\"'
service_rt = SshCommand.ssh_exec(username, node_ip, cmd)
instance_list = get_service_list()
if service_rt is not None:
try:
for svc in instance_list:
find_flag = False
for line in service_rt.splitlines():
line = line.decode('utf-8')
if line.startswith('ID'):
continue
id, name, mode, rep, img = line.split()
if svc == name:
find_flag = True
rep_tmp = rep.split('/')
if not (rep_tmp[0] == rep_tmp[1]):
service_status = 'nok'
svc_json = {
'name': name,
'mode': mode,
'replicas': rep,
'image': img,
'status': 'nok',
'monitor_item': True
}
fail_reason.append(svc_json)
else:
svc_json = {
'name': name,
'mode': mode,
'replicas': rep,
'image': img,
'status': 'ok',
'monitor_item': True
}
service_list.append(svc_json)
if not find_flag:
service_status = 'nok'
fail_reason.append('swarm ' + svc +
' service does not exist.')
break
for line in service_rt.splitlines():
line = line.decode('utf-8')
if line.startswith('ID'):
continue
id, name, mode, rep, img = line.split()
if name in instance_list:
continue
rep_tmp = rep.split('/')
if not (rep_tmp[0] == rep_tmp[1]):
svc_json = {
'name': name,
'mode': mode,
'replicas': rep,
'image': img,
'status': 'nok',
'monitor_item': False
}
else:
svc_json = {
'name': name,
'mode': mode,
'replicas': rep,
'image': img,
'status': 'ok',
'monitor_item': False
}
service_list.append(svc_json)
except:
LOG.exception()
service_status = 'fail'
else:
LOG.error("\'%s\' Swarm Service Check Error", node_ip)
service_status = 'fail'
for app in instance_list:
cmd = 'ssh root@' + swarm_manager + ' \"sudo docker service ps ' + app + '\"'
ps_rt = SshCommand.ssh_exec(username, node_ip, cmd)
if ps_rt is not None:
for line in ps_rt.splitlines():
line = line.decode('utf-8')
if line.startswith('ID'):
continue
line = line.replace(' \_ ', '')
line = " ".join(line.split())
tmp = line.split(' ')
ps_json = {
'name': tmp[1],
'image': tmp[2],
'node': tmp[3],
'desired_state': tmp[4],
'current_state': tmp[5]
}
ps_list.append(ps_json)
else:
LOG.error("\'%s\' Swarm PS Check Error", node_ip)
try:
sql = 'UPDATE ' + DB.SWARM_TBL + \
' SET service = \"' + str(service_list) + '\",' + \
' ps = \"' + str(ps_list) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE SWARM SERVICE/PS INFO -----\n' +
sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] SWARM SERVICE/PS DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
service_status = 'fail'
return service_status, fail_reason
def onos_app_check(conn, db_log, node_name, node_ip):
try:
app_rt = SshCommand.onos_ssh_exec(node_ip, 'apps -a -s')
status = 'ok'
app_active_list = list()
app_list = []
fail_reason = []
if app_rt is not None:
for line in app_rt.splitlines():
app_active_list.append(line.split(".")[2].split()[0])
if not 'cpman' in app_active_list:
# activate cpman
LOG.info('Cpman does not exist. Activate cpman')
SshCommand.onos_ssh_exec(node_ip,
'app activate org.onosproject.cpman')
for app in CONF.onos()['app_list']:
if app in app_active_list:
app_json = {
'name': app,
'status': 'ok',
'monitor_item': True
}
app_active_list.remove(app)
else:
status = 'nok'
app_json = {
'name': app,
'status': 'nok',
'monitor_item': True
}
fail_reason.append(app_json)
app_list.append(app_json)
for app in app_active_list:
app_json = {'name': app, 'status': 'ok', 'monitor_item': False}
app_list.append(app_json)
else:
LOG.error("\'%s\' ONOS Application Check Error", node_ip)
status = 'fail'
app_list = 'fail'
try:
sql = 'UPDATE ' + DB.ONOS_TBL + \
' SET applist = \"' + str(app_list) + '\"' +\
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE ONOS APP INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] ONOS APP DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
status = 'fail'
return status, fail_reason
def periodic(conn, pre_stat, db_log):
try:
cur_info = {}
#LOG.info('Periodic checking %s', str(CONF.watchdog()['check_system']))
try:
node_list = cmd_proc.get_node_list('all', 'nodename, ip_addr, username, type, sub_type')
if not node_list:
LOG.info("Not Exist Node data ...")
return
except:
LOG.exception()
return
# Read cur alarm status
sql = 'SELECT nodename, item, grade FROM ' + DB.EVENT_TBL
db_log.write_log(sql)
cur_grade = conn.cursor().execute(sql).fetchall()
old_nok_count = 0;
for nodename, item, grade in cur_grade:
if not cur_info.has_key(nodename):
cur_info[nodename] = {}
cur_info[nodename][item] = grade
if grade != 'ok':
old_nok_count += 1
new_nok_count = 0;
for node_name, node_ip, user_name, type, sub_type in node_list:
#LOG.info('------------------------------------ ' + node_name + ' START ------------------------------------')
onos_cluster = 'fail'
onos_device = 'fail'
onos_link = 'fail'
onos_app = 'fail'
# ping check
ping = net_check(node_ip)
ping_reason = []
if ping != 'ok':
reason.append('ping check failed on ' + node_ip)
new_nok_count += 1
ping = alarm_event.process_event(conn, db_log, node_name, type, 'PING', cur_info[node_name]['PING'], ping, ping_reason)
if ping == 'ok':
if type.upper() == 'ONOS':
# check connection
onos_cluster, onos_device, onos_link, onos_app, cluster_reason, device_reason, link_reason, app_reason = chk_onos.onos_check(conn, db_log, node_name, node_ip)
onos_cluster = alarm_event.process_event(conn, db_log, node_name, type, 'ONOS_CLUSTER',
cur_info[node_name]['ONOS_CLUSTER'], onos_cluster, cluster_reason)
onos_device = alarm_event.process_event(conn, db_log, node_name, type, 'ONOS_DEVICE',
cur_info[node_name]['ONOS_DEVICE'], onos_device, device_reason)
onos_link = alarm_event.process_event(conn, db_log, node_name, type, 'ONOS_LINK',
cur_info[node_name]['ONOS_LINK'], onos_link, link_reason)
onos_app = alarm_event.process_event(conn, db_log, node_name, type, 'ONOS_APP',
cur_info[node_name]['ONOS_APP'], onos_app, app_reason)
if onos_cluster != 'ok': new_nok_count += 1
if onos_device != 'ok': new_nok_count += 1
if onos_link != 'ok': new_nok_count += 1
if onos_app != 'ok': new_nok_count += 1
try:
sql = 'UPDATE ' + DB.STATUS_TBL + \
' SET' + \
' PING = \'' + ping + '\',' + \
' ONOS_CLUSTER = \'' + onos_cluster + '\',' + \
' ONOS_DEVICE = \'' + onos_device + '\',' + \
' ONOS_LINK = \'' + onos_link + '\',' + \
' ONOS_APP = \'' + onos_app + '\',' + \
' time = \'' + str(datetime.now()) + '\'' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE TOTAL SYSTEM INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] TOTAL SYSTEM INFO DB Update Fail.')
except:
LOG.exception()
# do not version log on everthing is ok
if old_nok_count > 0:
LOG.info('chk_onos[%s]: ping=%s cluster=%s device=%s link=%s app=%s' %
(node_name, ping, onos_cluster, onos_device, onos_link, onos_app))
if old_nok_count > 0 and new_nok_count == 0:
alarm_event.process_event(conn, db_log, 'ALL', 'SITE', 'STATUS', 'none', 'ok', [])
# send all alarm messages pending
alarm_event.flush_event_alarm();
except:
LOG.exception()
return pre_stat
def vrouter_check(conn, db_log, node_name, user_name, node_ip):
ret_docker = 'ok'
docker_list = []
fail_list = []
onos_id = ''
docker_rt = SshCommand.ssh_exec(user_name, node_ip, 'sudo docker ps')
if docker_rt is not None:
try:
for docker in CONF.openstack()['docker_list']:
for line in docker_rt.splitlines():
if line.startswith('CONTAINER'):
continue
tmp_line = line.split()
if ' ' + docker in line:
if not 'Up' in line:
docker_json = {'name': docker, 'status': 'nok', 'type': 'docker'}
fail_list.append(docker_json)
ret_docker = 'nok'
else:
docker_json = {'name': docker, 'status': 'ok', 'type': 'docker'}
docker_list.append(docker_json)
if 'onos' in tmp_line[1]:
onos_id = tmp_line[0]
except:
LOG.exception()
else:
LOG.error("\'%s\' Vrouter Node Check Error", node_ip)
ret_docker = 'fail'
onos_app_list = []
route_list = []
if not onos_id == '':
try:
# get onos container ip
onos_rt = SshCommand.ssh_exec(user_name, node_ip, 'sudo docker inspect ' + onos_id + ' | grep IPAddress')
if onos_rt is not None:
for line in onos_rt.splitlines():
line = line.strip()
if line.startswith('\"IPAddress'):
tmp = line.split(':')
onos_ip = tmp[1].strip().replace('\"', '').replace(',', '')
break
app_list = SshCommand.ssh_pexpect(user_name, node_ip, onos_ip, 'apps -a -s')
app_active_list = list()
for line in app_list.splitlines():
if line.startswith('fail'):
continue
app_active_list.append(line.split(".")[2].split()[0])
for app in CONF.openstack()['onos_vrouter_app_list']:
if app in app_active_list:
app_json = {'name': app, 'status': 'ok', 'type': 'onos_app'}
else:
app_json = {'name': app, 'status': 'nok', 'type': 'onos_app'}
fail_list.append(app_json)
ret_docker = 'nok'
onos_app_list.append(app_json)
str_route = SshCommand.ssh_pexpect(user_name, node_ip, onos_ip, 'routes')
for line in str_route.splitlines():
line = line.strip()
if (line.startswith('Table') or line.startswith('Network') or line.startswith('Total')):
continue
new_line = " ".join(line.split())
if new_line.startswith('fail'):
continue
tmp = new_line.split(' ')
route_json = {'network': tmp[0], 'next_hop': tmp[1]}
route_list.append(route_json)
except:
LOG.exception()
else:
LOG.info('can not find onos_id.')
ret_docker = 'fail'
try:
sql = 'UPDATE ' + DB.OPENSTACK_TBL + \
' SET docker = \"' + str(docker_list) + '\",' + \
' onosApp = \"' + str(onos_app_list) + '\",' + \
' routingTable = \"' + str(route_list) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE GATEWAY INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] GATEWAY DB Update Fail.')
except:
LOG.exception()
return ret_docker, fail_list
def run(self):
db_log = USER_LOG()
db_log.set_log('db.log',
CONF.base()['log_rotate_time'],
CONF.base()['log_backup_count'])
pre_stat = dict()
# DB initiation
DB.db_initiation(db_log)
# Start RESTful server
try:
REST_SVR.rest_server_start()
except:
print 'Rest Server failed to start'
LOG.exception()
self.exit()
# Periodic monitoring
if CONF.watchdog()['interval'] == 0:
LOG.info("--- Not running periodic monitoring ---")
while True:
time.sleep(3600)
else:
LOG.info("--- Periodic Monitoring Start ---")
history_log.write_log("--- Event History Start ---")
conn = DB.connection()
exitFlag = False
while True:
try:
i = 0
while i < 3:
i = i + 1
# check rest server
try:
url = 'http://' + socket.gethostbyname(
socket.gethostname()) + ':' + str(CONF.rest(
)['rest_server_port']) + '/alive-check'
cmd = 'curl -X GET \"' + url + '\"'
LOG.info('cmd = ' + cmd)
result = Popen(cmd,
stdout=PIPE,
stderr=PIPE,
shell=True)
output, error = result.communicate()
if result.returncode != 0:
LOG.info('REST SERVER CHECK FAIL [' + str(i) +
']')
if i == 3:
LOG.info('fail to check rest server.')
alarm_event.push_event(
'sonawatcher',
'SONAWATCHER_DISCONNECT', 'critical',
'normal',
'sonawatcher server shutdown',
str(datetime.now()))
conn.close()
exitFlag = True
self.exit()
break
else:
break
except:
LOG.exception()
if exitFlag:
break
pre_stat = watchdog.periodic(conn, pre_stat, db_log)
time.sleep(CONF.watchdog()['interval'])
except:
alarm_event.push_event('sonawatcher',
'SONAWATCHER_DISCONNECT',
'critical', 'normal',
'sonawatcher server shutdown',
str(datetime.now()))
conn.close()
LOG.exception()
def get_node_traffic(conn, db_log, node_name, rx_dic, tx_dic, total_rx, total_tx, err_info, pre_stat):
try:
status = 'ok'
reason_list = []
pre_total_rx = total_rx
pre_total_tx = total_tx
# check minimum packet count
sql = 'SELECT data_ip FROM ' + DB.OPENSTACK_TBL + ' WHERE nodename = \'' + node_name + '\''
data_ip = conn.cursor().execute(sql).fetchone()[0]
sql = 'SELECT ip_addr FROM ' + DB.NODE_INFO_TBL + ' WHERE type = \'ONOS\''
nodes_info = conn.cursor().execute(sql).fetchall()
min_rx = 0
if len(nodes_info) == 0:
LOG.info('Fail to load onos list')
status = 'fail'
else:
for ip in nodes_info:
flows_rt = SshCommand.onos_ssh_exec(ip[0], '\"flows --filter \'{tunnelDst=' + data_ip + '}\' --short\"')
if flows_rt is not None:
for line in flows_rt.splitlines():
if 'tunnelDst' in line:
min_rx = min_rx + int(line.split(',')[2].split('=')[1])
break
if not dict(pre_stat).has_key(node_name + '_VXLAN'):
status = '-'
ratio = -1
else:
total_rx = total_rx - int(dict(pre_stat)[node_name + '_VXLAN']['total_rx'])
total_tx = total_tx - int(dict(pre_stat)[node_name + '_VXLAN']['total_tx'])
cur_min = min_rx - int(dict(pre_stat)[node_name + '_VXLAN']['min_rx'])
if total_rx == 0 and total_tx == 0:
ratio = 100
elif total_tx <= 0 or total_tx < 0:
LOG.info('Node Traffic Ratio Fail.')
ratio = 0
else:
ratio = float(total_rx) * 100 / total_tx
LOG.info('Node Traffic Ratio = ' + str(ratio))
port_json = {'rx': rx_dic[node_name], 'minimum_rx': min_rx, 'rx_drop': err_info['rx_drop'], 'rx_errs': err_info['rx_err'],
'tx': tx_dic[node_name], 'tx_drop': err_info['tx_drop'], 'tx_errs': err_info['tx_err']}
description = ''
if not status == '-':
description = 'Ratio of success for all nodes = ' + str(ratio) + ' (' + str(total_rx) + ' / ' + str(total_tx) + ')'
if ratio < float(CONF.alarm()['node_traffic_ratio']):
LOG.info('[NODE TRAFFIC] ratio nok')
status = 'nok'
if total_rx < cur_min:
LOG.info('CUR_MIN_RX = ' + str(cur_min) + ', CUR_RX = ' + str(total_rx) + ', Less than rx minimum.')
status = 'nok'
if err_info['rx_drop'] - int(dict(pre_stat)[node_name + '_VXLAN']['rx_drop']) > 0:
LOG.info('[NODE TRAFFIC] rx_drop nok')
status = 'nok'
if err_info['rx_err'] - int(dict(pre_stat)[node_name + '_VXLAN']['rx_err']) > 0:
LOG.info('[NODE TRAFFIC] rx_err nok')
status = 'nok'
if err_info['tx_drop'] - int(dict(pre_stat)[node_name + '_VXLAN']['tx_drop']) > 0:
LOG.info('[NODE TRAFFIC] tx_drop nok')
status = 'nok'
if err_info['tx_err'] - int(dict(pre_stat)[node_name + '_VXLAN']['tx_err']) > 0:
LOG.info('[NODE TRAFFIC] tx_err nok')
status = 'nok'
in_out_dic = dict()
in_out_dic['total_rx'] = pre_total_rx
in_out_dic['total_tx'] = pre_total_tx
in_out_dic['min_rx'] = min_rx
in_out_dic['rx_drop'] = err_info['rx_drop']
in_out_dic['rx_err'] = err_info['rx_err']
in_out_dic['tx_drop'] = err_info['tx_drop']
in_out_dic['tx_err'] = err_info['tx_err']
pre_stat[node_name + '_VXLAN'] = in_out_dic
except:
LOG.exception()
status = 'fail'
vxlan_json = {'port_stat_vxlan': port_json, 'period': CONF.watchdog()['interval'],
'ratio': format(ratio, '.2f'), 'current_rx': total_rx, 'current_tx': total_tx,
'description': description, 'threshold': CONF.alarm()['node_traffic_ratio'], 'status': status}
try:
sql = 'UPDATE ' + DB.OPENSTACK_TBL + \
' SET vxlan_traffic = \"' + str(vxlan_json) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE VXLAN STAT INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] VXLAN STAT DB Update Fail.')
except:
LOG.exception()
if not status == 'ok':
reason_list.append(vxlan_json)
return status, pre_stat, reason_list
def get_internal_traffic(conn, db_log, node_name, node_ip, user_name, sub_type, rx_count, patch_tx, pre_stat):
try:
status = 'ok'
in_packet = 0
out_packet = 0
reason_list = []
desc = ''
if sub_type == 'COMPUTE':
flow_rt = SshCommand.ssh_exec(user_name, node_ip, 'sudo ovs-ofctl -O OpenFlow13 dump-flows br-int')
inport_cnt = 0
gw_cnt = 0
output_cnt = 0
if flow_rt is not None:
for line in flow_rt.splitlines():
tmp = line.split(',')
if 'in_port' in line:
inport_cnt = inport_cnt + int(tmp[3].split('=')[1])
elif 'output' in line:
output_cnt = output_cnt + int(tmp[3].split('=')[1])
elif 'actions=group' in line:
gw_cnt = gw_cnt + int(tmp[3].split('=')[1])
in_packet = inport_cnt + rx_count
out_packet = gw_cnt + output_cnt
port_json = {'vm_tx': inport_cnt, 'vxlan_rx': rx_count, 'out_gw': gw_cnt, 'output': output_cnt}
else:
port_json = {'vm_tx': -1, 'vxlan_rx': -1, 'out_gw': -1, 'output': -1}
status = 'fail'
else:
port_json = {'vxlan_rx': rx_count, 'patch-integ': patch_tx}
if patch_tx == -1:
status = 'fail'
else:
in_packet = rx_count
out_packet = patch_tx
for_save_in = in_packet
for_save_out = out_packet
if not dict(pre_stat).has_key(node_name + '_internal'):
status = '-'
vxlan_json = {'port_stat_in_out': port_json, 'period': CONF.watchdog()['interval'],
'ratio': 0, 'current_rx': -1, 'current_tx': -1,
'description': desc, 'threshold': CONF.alarm()['internal_traffic_ratio'], 'status': status}
elif status == 'ok':
in_packet = in_packet - int(dict(pre_stat)[node_name + '_internal']['in_packet'])
out_packet = out_packet - int(dict(pre_stat)[node_name + '_internal']['out_packet'])
if in_packet == 0 and out_packet == 0:
ratio = 100
elif in_packet <= 0 or out_packet < 0:
LOG.info('Internal Traffic Ratio Fail.')
ratio = 0
else:
ratio = float(out_packet) * 100 / in_packet
LOG.info('Internal Traffic Ratio = ' + str(ratio))
desc = 'Internal Traffic Ratio = ' + str(ratio) + '(' + str(out_packet) + '/' + str(in_packet) + ')'
if ratio < float(CONF.alarm()['internal_traffic_ratio']):
status = 'nok'
vxlan_json = {'port_stat_in_out': port_json, 'period': CONF.watchdog()['interval'],
'ratio': format(ratio, '.2f'), 'current_rx': in_packet, 'current_tx': out_packet,
'description': desc, 'threshold': CONF.alarm()['internal_traffic_ratio'], 'status': status}
in_out_dic = dict()
in_out_dic['in_packet'] = for_save_in
in_out_dic['out_packet'] = for_save_out
pre_stat[node_name + '_internal'] = in_out_dic
try:
sql = 'UPDATE ' + DB.OPENSTACK_TBL + \
' SET internal_traffic = \"' + str(vxlan_json) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE INTERNAL TRAFFIC INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] INTERNAL TRAFFIC DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
status = 'fail'
if not status == 'ok':
reason_list.append(vxlan_json)
return status, pre_stat, reason_list
def get_gw_ratio_compute(conn, db_log, node_ip, node_name, pre_stat):
status = 'ok'
reason = []
try:
sql = 'SELECT ' + DB.ONOS_TBL + '.nodename, nodelist, ip_addr' + ' FROM ' + DB.ONOS_TBL + \
' INNER JOIN ' + DB.NODE_INFO_TBL + ' ON ' + DB.ONOS_TBL + '.nodename = ' + DB.NODE_INFO_TBL + '.nodename'
nodes_info = conn.cursor().execute(sql).fetchall()
if len(nodes_info) == 0:
LOG.info('Fail to load onos list')
return 'fail', pre_stat, reason
manage_ip = ''
hostname = ''
for nodename, nodelist, ip in nodes_info:
if not nodelist == 'none':
for node_info in eval(nodelist):
try:
if dict(node_info)['management_ip'] == node_ip:
manage_ip = ip
hostname = dict(node_info)['hostname']
except:
manage_ip = ''
if not manage_ip == '':
break
if not manage_ip == '':
break
if hostname == '':
LOG.info('Can not find hostname')
return 'fail', pre_stat, reason
try:
sql = 'SELECT of_id FROM ' + DB.OPENSTACK_TBL + ' WHERE hostname = \'' + str(hostname) + '\''
LOG.info(sql)
node_info = conn.cursor().execute(sql).fetchone()
of_id = node_info[0]
except:
LOG.exception()
LOG.info('Can not find of_id')
return 'fail', pre_stat, reason
group_rt = SshCommand.onos_ssh_exec(manage_ip, 'groups')
total_cnt = 0
gw_list = []
if group_rt is not None:
for line in group_rt.splitlines():
if of_id in line:
tmp = line.split(',')
for col in tmp:
if 'packets=' in col:
total_cnt = total_cnt + int(col.split('=')[1])
gw_list.append(int(col.split('=')[1]))
str_ratio = ''
if not dict(pre_stat).has_key(node_name + '_GW'):
status = '-'
json_ratio = {'ratio': '-', 'status': status, 'period':CONF.watchdog()['interval'], 'status': status}
else:
i = 0
for gw in gw_list:
cur_gw = gw - pre_stat[node_name + '_GW']['gw_list'][i]
cur_total = total_cnt - pre_stat[node_name + '_GW']['gw_total']
LOG.info('cur_gw = ' + str(cur_gw))
LOG.info('cur_total = ' + str(cur_total))
if cur_gw == 0 and cur_total == 0:
ratio = 100/len(gw_list)
elif cur_gw <= 0 or cur_total <= 0:
ratio = 0
else:
ratio = float(cur_gw) * 100 / cur_total
i = i + 1
str_ratio = str_ratio + str(ratio) + ':'
if ratio < float(CONF.alarm()['gw_ratio']):
status = 'nok'
json_ratio = {'ratio': str_ratio.rstrip(':'), 'status': status, 'period':CONF.watchdog()['interval'], 'status': status}
LOG.info('[COMPUTE] ' + 'GW_RATIO = ' + str_ratio.rstrip(':'))
try:
sql = 'UPDATE ' + DB.OPENSTACK_TBL + \
' SET gw_ratio = \"' + str(json_ratio) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE TRAFFIC GW INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] TRAFFIC GW DB Update Fail.')
except:
LOG.exception()
in_out_dic = dict()
in_out_dic['gw_list'] = gw_list
in_out_dic['gw_total'] = total_cnt
pre_stat[node_name + '_GW'] = in_out_dic
except:
LOG.exception()
status = 'fail'
if not status == 'ok':
reason.append(json_ratio)
return status, pre_stat, reason
def get_gw_ratio_gateway(conn, db_log, node_ip, node_name, rx, gw_rx_sum, pre_stat):
status = 'ok'
reason = []
try:
sql = 'SELECT ' + DB.ONOS_TBL + '.nodename, nodelist, ip_addr' + ' FROM ' + DB.ONOS_TBL + \
' INNER JOIN ' + DB.NODE_INFO_TBL + ' ON ' + DB.ONOS_TBL + '.nodename = ' + DB.NODE_INFO_TBL + '.nodename'
nodes_info = conn.cursor().execute(sql).fetchall()
if len(nodes_info) == 0:
LOG.info('Fail to load onos list')
return 'fail', pre_stat, reason
# search data_ip
data_ip = ''
manage_ip = ''
cpt_to_gw_packet = 0
for nodename, nodelist, ip in nodes_info:
if not nodelist == 'none':
for node_info in eval(nodelist):
try:
if dict(node_info)['management_ip'] == node_ip:
manage_ip = ip
data_ip = dict(node_info)['data_ip']
except:
manage_ip = ''
if not manage_ip == '':
break
if not manage_ip == '':
break
if data_ip == '':
LOG.info('Can not find data ip')
return 'fail', pre_stat, reason
group_rt = SshCommand.onos_ssh_exec(manage_ip, 'groups')
if group_rt is not None:
for line in group_rt.splitlines():
if '{tunnelDst=' + data_ip + '}' in line:
tmp = line.split(',')
for col in tmp:
if 'packets=' in col:
cpt_to_gw_packet = cpt_to_gw_packet + int(col.split('=')[1])
if not dict(pre_stat).has_key(node_name + '_GW'):
status = '-'
json_ratio = {'current_rx': '-', 'current_compute_tx': '-', 'current_total': '-',
'ratio': '-',
'period': CONF.watchdog()['interval'], 'status': status, 'packet_loss': False,
'description': ''}
else:
cur_rx = rx - int(dict(pre_stat)[node_name + '_GW']['rx'])
cur_total = gw_rx_sum - int(dict(pre_stat)[node_name + '_GW']['gw_rx_sum'])
cur_packet = cpt_to_gw_packet - int(dict(pre_stat)[node_name + '_GW']['cpt_to_gw_packet'])
if cur_rx == 0 and cur_total == 0:
ratio = 100
elif cur_rx <= 0 or cur_total < 0:
ratio = 0
else:
ratio = float(cur_rx) * 100 / cur_total
desc = 'GW RATIO = ' + str(ratio) + ' (' + str(cur_rx) + ' / ' + str(cur_total) + ')'
loss_flag = False
if cur_rx < cur_packet:
LOG.info('GW Ratio Fail. (Data loss)')
loss_flag = True
LOG.info('GW Ratio = ' + str(ratio))
if ratio < float(CONF.alarm()['gw_ratio']) or cur_rx < cur_packet:
status = 'nok'
json_ratio = {'current_rx': cur_rx, 'current_compute_tx': cur_packet, 'current_total': cur_total, 'ratio': format(ratio, '.2f'),
'period':CONF.watchdog()['interval'], 'status': status, 'packet_loss': loss_flag, 'description': desc}
try:
sql = 'UPDATE ' + DB.OPENSTACK_TBL + \
' SET gw_ratio = \"' + str(json_ratio) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE TRAFFIC GW INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] TRAFFIC GW DB Update Fail.')
except:
LOG.exception()
in_out_dic = dict()
in_out_dic['rx'] = rx
in_out_dic['gw_rx_sum'] = gw_rx_sum
in_out_dic['cpt_to_gw_packet'] = cpt_to_gw_packet
pre_stat[node_name + '_GW'] = in_out_dic
except:
LOG.exception()
status = 'fail'
if not status == 'ok':
reason.append(json_ratio)
return status, pre_stat, reason
def onos_rest_check(conn, db_log, node_name, node_ip):
try:
web_status = 'ok'
web_list = []
fail_reason = []
web_rt = SshCommand.onos_ssh_exec(node_ip, 'web:list')
if web_rt is not None:
for web in CONF.onos()['rest_list']:
for line in web_rt.splitlines():
if line.startswith('ID') or line.startswith('--'):
continue
if ' ' + web + ' ' in line:
if not ('Active' in line and 'Deployed' in line):
rest_json = {
'name': web,
'status': 'nok',
'monitor_item': True
}
fail_reason.append(rest_json)
web_status = 'nok'
else:
rest_json = {
'name': web,
'status': 'ok',
'monitor_item': True
}
web_list.append(rest_json)
for line in web_rt.splitlines():
if line.startswith('ID') or line.startswith('--'):
continue
name = " ".join(line.split()).split(' ')[10]
if not name in CONF.onos()['rest_list']:
if not ('Active' in line and 'Deployed' in line):
rest_json = {
'name': name,
'status': 'nok',
'monitor_item': False
}
else:
rest_json = {
'name': name,
'status': 'ok',
'monitor_item': False
}
web_list.append(rest_json)
else:
LOG.error("\'%s\' ONOS Rest Check Error", node_ip)
web_status = 'fail'
web_list = 'fail'
try:
sql = 'UPDATE ' + DB.ONOS_TBL + \
' SET weblist = \"' + str(web_list) + '\"' +\
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE ONOS REST INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] ONOS REST DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
web_status = 'fail'
return web_status, fail_reason
def xos_sync_check(conn, db_log, node_name):
swarm_sync = 'ok'
sync_list = []
fail_reason = []
try:
url = CONF.xos()['xos_rest_server']
account = CONF.xos()['xos_rest_account']
cmd = 'curl -H "Accept: application/json; indent=4" -u ' + account + ' -X GET ' + url + '/api/core/diags/'
result = Popen(cmd, stdout=PIPE, stderr=PIPE, shell=True)
output, error = result.communicate()
if result.returncode != 0:
LOG.error("Cmd Fail, cause => %s", error)
return 'fail', None
sync_array = json.loads(output)
for xos_info in sync_array:
backend_status = xos_info['backend_status']
LOG.info('xos_sync_backend_status = ' + backend_status)
tmp = str(backend_status).split('-')
if tmp[0].strip() in ['0', '1']:
status = 'ok'
else:
status = 'nok'
# check time
last_time = json.loads(xos_info['backend_register'])['last_run']
cur_time = time.time()
interval = cur_time - last_time
interval = int(interval)
if interval >= 30:
status = 'nok'
xos_json = {
'name': xos_info['name'],
'status': status,
'description': tmp[1].strip(),
'last_run_interval': interval
}
sync_list.append(xos_json)
if status == 'nok':
swarm_sync = 'nok'
fail_reason.append(xos_json)
try:
sql = 'UPDATE ' + DB.XOS_TBL + \
' SET synchronizer = \"' + str(sync_list) + '\"' + \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE SYNCHRONIZER INFO -----\n' +
sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] SYNCHRONIZER DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
swarm_sync = 'fail'
return swarm_sync, fail_reason
def onos_check(conn, db_log, node_name, node_ip):
# called on each ONOS node in NODE_INFO_TBL
try:
# check cluster nodes
node_list = []
node_status = 'ok'
node_fail_reason = []
ret, rsp = onos_api_req(node_ip, 'onos/v1/cluster')
if rsp is not None:
try:
node_tbl = dict()
for node in rsp['nodes']:
node_tbl[node['ip']] = node
for onos_node in CONF.onos()['list']:
if len(onos_node.split(':')) != 2:
continue
id = onos_node.split(':')[0]
ip = onos_node.split(':')[1]
if id is '' or ip is '':
continue
if ip in node_tbl:
node = node_tbl[ip]
node['id'] = id
node['monitor_item'] = True
if node['status'] != 'READY':
node_status = 'nok'
node_fail_reason.append('Node ' + id + ' DOWN')
node_tbl.pop(ip)
else:
node = {
'id': id,
'ip': ip,
'status': 'nok',
'monitor_item': True
}
node_status = 'nok'
node_fail_reason.append('Node ' + id + ' DOWN')
node_list.append(node)
for node in node_tbl.values():
node['monitor_item'] = False
node_list.append(node)
except:
LOG.exception()
LOG.error("\'%s\' ONOS Check Error(nodes)", node_ip)
node_status = 'fail'
# check devices
device_list = []
device_status = 'ok'
device_fail_reason = []
ret, rsp = onos_api_req(node_ip, 'onos/v1/devices')
if rsp is not None:
try:
device_tbl = dict()
for device in rsp['devices']:
device['id'] = 'of:' + device['chassisId'].rjust(16, '0')
device_tbl[device['id']] = device
for id in CONF.onos()['device_list']:
if id is '':
continue
# no config
if id in device_tbl:
device = device_tbl[id]
device['monitor_item'] = True
if not device['available']:
device_status = 'nok'
device_fail_reason.append('Device ' + id + ' DOWN')
device_tbl.pop(id)
else:
device = {
'id': id,
'available': False,
'channelId': '-',
'name': '-',
'role': '-',
'monitor_item': True
}
device_status = 'nok'
device_fail_reason.append('Device ' + id + ' DOWN')
device_list.append(device)
for device in device_tbl.values():
device['monitor_item'] = False
device_list.append(device)
except:
LOG.exception()
LOG.error("\'%s\' ONOS Check Error(devices)", node_ip)
device_status = 'fail'
else:
LOG.error("\'%s\' ONOS Check Error(devices)", node_ip)
device_status = 'fail'
# check links
link_list = []
link_status = 'ok'
link_fail_reason = []
ret, rsp = onos_api_req(node_ip, 'onos/v1/links')
if rsp is not None:
try:
link_tbl = dict()
for link in rsp['links']:
link['src'] = link['src']['device'] + '/' + link['src'][
'port']
link['dst'] = link['dst']['device'] + '/' + link['dst'][
'port']
link_tbl[link['src'] + '-' + link['dst']] = link
for id in CONF.onos()['link_list']:
if id is '':
continue
if len(id.split('-')) != 2:
link = {
'src': id,
'dst': '(invalid_link_config)',
'expected': 'false',
'state': '-',
'type': "-",
'monitor_item': True
}
link_status = 'nok'
link_fail_reason.append(
'Link ' + id + ' is configed as INVALID ID FORMAT')
link_list.append(link)
continue
if id in link_tbl:
link = link_tbl[id]
link['monitor_item'] = True
if link['state'] != 'ACTIVE':
link_status = 'nok'
link_fail_reason.append('Link ' + id + ' DOWN')
link_list.append(link)
link_tbl.pop(id)
else:
link = {
'src': id.split('-')[0],
'dst': id.split('-')[1],
'expected': 'false',
'state': '-',
'type': "-",
'monitor_item': True
}
link_status = 'nok'
link_fail_reason.append('Link ' + id + ' DOWN')
link_list.append(link)
rev_id = id.split('-')[1] + '-' + id.split('-')[0]
if rev_id in link_tbl:
link = link_tbl[rev_id]
link['monitor_item'] = True
if link['state'] != 'ACTIVE':
link_status = 'nok'
link_fail_reason.append('Link' + rev_id + ' DOWN')
link_list.append(link)
link_tbl.pop(rev_id)
else:
link = {
'src': rev_id.split('-')[0],
'dst': rev_id.split('-')[1],
'expected': 'false',
'state': '-',
'type': "-",
'monitor_item': True
}
link_status = 'nok'
link_fail_reason.append('Link ' + rev_id + ' DOWN')
link_list.append(link)
for link in link_tbl.values():
link['monitor_item'] = False
link_list.append(link)
except:
LOG.exception()
LOG.error("\'%s\' ONOS Check Error(links)", node_ip)
link_status = 'fail'
# check apps
app_list = []
app_status = 'ok'
app_fail_reason = []
ret, rsp = onos_api_req(node_ip, 'onos/v1/applications')
if rsp is not None:
try:
active_app_list = []
for app_rsp in rsp['applications']:
if app_rsp['state'] == 'ACTIVE':
active_app_list.append(app_rsp['name'].replace(
'org.onosproject.', ''))
for app in CONF.onos()['app_list']:
if app in active_app_list:
app_json = {
'name': app,
'status': 'ok',
'monitor_item': True
}
active_app_list.remove(app)
else:
app_json = {
'name': app,
'status': 'nok',
'monitor_item': True
}
app_status = 'nok'
app_fail_reason.append(app_json)
app_list.append(app_json)
for app in active_app_list:
app_json = {
'name': app,
'status': 'ok',
'monitor_item': False
}
app_list.append(app_json)
except:
LOG.exception()
LOG.error("\'%s\' ONOS Check Error(apps)", node_ip)
app_status = 'fail'
else:
LOG.error("\'%s\' ONOS Check Error(apps)", node_ip)
link_status = 'fail'
# store to db
try:
sql = 'UPDATE ' + DB.ONOS_TBL + \
' SET ' + \
' cluster = \"' + str(node_list) + '\",' \
' device = \"' + str(device_list) + '\",' \
' link = \"' + str(link_list) + '\",' \
' app = \"' + str(app_list) + '\"' \
' WHERE nodename = \'' + node_name + '\''
db_log.write_log('----- UPDATE ONOS CONNECTION INFO -----\n' + sql)
if DB.sql_execute(sql, conn) != 'SUCCESS':
db_log.write_log('[FAIL] ONOS CONNECTION DB Update Fail.')
except:
LOG.exception()
except:
LOG.exception()
cluster_status = 'fail'
device_status = 'fail'
link_status = 'fail'
app_status = 'fail'
return node_status, device_status, link_status, app_status, node_fail_reason, device_fail_reason, link_fail_reason, app_fail_reason