def module_init(args): global qh comp = cconf.parse_conf(nagios_cfg) for v in comp.params: if v[0] == 'query_socket': qh = v[1] break
def module_init(args):
global qh
global single
rem_args = []
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
for arg in args:
if arg.startswith('--socket='):
qh = arg.split('=', 1)[1]
continue
if arg == '--single':
single = True
continue
rem_args.append(arg)
if not qh:
if os.access(nagios_cfg, os.R_OK):
comp = cconf.parse_nagios_cfg(nagios_cfg)
qh = comp.query_socket
else:
print "ERROR: Unable to find Naemon query socket"
return 1
return rem_args
def module_init(args):
global qh
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
def cmd_spool(args=False):
"""[--maxage=<seconds>] [--warning=X] [--critical=X] <path> [--delete]
Checks a certain spool directory for files (and files only) that
are older than 'maxage'. It is intended to prevent buildup of
checkresult files and unprocessed performance-data files in the
various spool directories used by op5 Monitor.
--delete causes too old files to be removed.
--maxage is given in seconds and defaults to 300 (5 minutes).
<path> may be 'perfdata' or 'checks', in which case directory
names will be taken from op5 defaults.
--warning and --critical have no effect if '--delete' is given
and will otherwise specify threshold values.
Only one directory at a time may be checked.
"""
maxage = 300
warning = 5
critical = 10
path = False
delete = False
npcd_config = '/opt/monitor/etc/pnp/npcd.cfg'
for arg in args:
if arg.startswith('--maxage='):
maxage = str_to_seconds(arg.split('=', 1)[1])
elif arg.startswith('--warning='):
warning = int(arg.split('=', 1)[1])
elif arg.startswith('--critical='):
critical = int(arg.split('=', 1)[1])
elif arg == '--delete':
delete = True
elif path == False:
path = arg
if path == False:
nplug.unknown("'path' is a required argument")
if path == 'checks':
path = check_result_path
elif path == 'perfdata':
if os.access(npcd_config, os.R_OK):
comp = cconf.parse_conf(npcd_config)
for k, v in comp.params:
if k == 'perfdata_spool_dir':
path = v
break
comp = False
else:
path = '/opt/monitor/var/spool/perfdata'
bad = 0
bad_paths = []
now = int(time.time())
try:
result = get_files(path)
except OSError, e:
nplug.die(nplug.STATE_UNKNOWN,
"Spool directory \"%s\" doesn't exist" % (path, ))
def cmd_spool(args=False):
"""[--maxage=<seconds>] [--warning=X] [--critical=X] <path> [--delete]
Checks a certain spool directory for files (and files only) that
are older than 'maxage'. It is intended to prevent buildup of
checkresult files and unprocessed performance-data files in the
various spool directories used by op5 Monitor.
--delete causes too old files to be removed.
--maxage is given in seconds and defaults to 300 (5 minutes).
<path> may be 'perfdata' or 'checks', in which case directory
names will be taken from op5 defaults.
--warning and --critical have no effect if '--delete' is given
and will otherwise specify threshold values.
Only one directory at a time may be checked.
"""
maxage = 300
warning = 5
critical = 10
path = False
delete = False
npcd_config = '/opt/monitor/etc/pnp/npcd.cfg'
for arg in args:
if arg.startswith('--maxage='):
maxage = str_to_seconds(arg.split('=', 1)[1])
elif arg.startswith('--warning='):
warning = int(arg.split('=', 1)[1])
elif arg.startswith('--critical='):
critical = int(arg.split('=', 1)[1])
elif arg == '--delete':
delete = True
elif path == False:
path = arg
if path == False:
nplug.unknown("'path' is a required argument")
if path == 'checks':
path = check_result_path
elif path == 'perfdata':
if os.access(npcd_config, os.R_OK):
comp = cconf.parse_conf(npcd_config)
for k, v in comp.params:
if k == 'perfdata_spool_dir':
path = v
break
comp = False
else:
path = '/opt/monitor/var/spool/perfdata'
bad = 0
bad_paths = []
now = int(time.time())
try:
result = get_files(path)
except OSError, e:
nplug.die(nplug.STATE_UNKNOWN, "Spool directory \"%s\" doesn't exist" % (path,))
def module_init(args):
global qh
rem_args = []
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
for arg in args:
if arg.startswith('--socket='):
qh = arg.split('=', 1)[1]
continue
rem_args.append(arg)
return rem_args
def module_init(args):
global qh
rem_args = []
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
for arg in args:
if arg.startswith('--socket='):
qh = arg.split('=', 1)[1]
continue
rem_args.append(arg)
return rem_args
def module_init(args):
global wanted_types, wanted_names
global have_type_arg, have_name_arg
global mst, lsc, ls_path, qh
rem_args = []
i = -1
for arg in args:
i += 1
if arg == '--':
rem_args += args[i:]
break
if arg.startswith('--merlin-cfg=') or arg.startswith('--config='):
mconf.config_file = arg.split('=', 1)[1]
elif arg.startswith('--type='):
wanted_types = arg.split('=')[1].split(',')
have_type_arg = True
elif arg.startswith('--name='):
wanted_names = arg.split('=')[1].split(',')
have_name_arg = True
else:
# not an argument we recognize, so stash it and move on
rem_args += [arg]
continue
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
elif v[0] == 'broker_module' and 'livestatus' in v[1]:
ary = v[1].rsplit(' ')
for p in ary[1:]:
if not '=' in p:
ls_path = p
break
lsc = livestatus.SingleSiteConnection('unix:' + ls_path)
mst = merlin_status(lsc, qh)
return rem_args
def module_init(args):
global wanted_types, wanted_names
global have_type_arg, have_name_arg
global mst, lsc, ls_path, qh
rem_args = []
i = -1
for arg in args:
i += 1
if arg == '--':
rem_args += args[i:]
break
if arg.startswith('--merlin-cfg=') or arg.startswith('--config='):
mconf.config_file = arg.split('=', 1)[1]
elif arg.startswith('--type='):
wanted_types = arg.split('=')[1].split(',')
have_type_arg = True
elif arg.startswith('--name='):
wanted_names = arg.split('=')[1].split(',')
have_name_arg = True
else:
# not an argument we recognize, so stash it and move on
rem_args += [arg]
continue
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
elif v[0] == 'broker_module' and 'livestatus' in v[1]:
ary = v[1].rsplit(' ')
for p in ary[1:]:
if not '=' in p:
ls_path = p
break
lsc = livestatus.SingleSiteConnection('unix:' + ls_path)
mst = merlin_status(lsc, qh)
return rem_args
def module_init(args):
global wanted_types, wanted_names
global have_type_arg, have_name_arg
global qh
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
wanted_types = mconf.merlin_node.valid_types
rem_args = []
i = -1
for arg in args:
i += 1
# stop parsing immediately if we hit double-dashes, or we
# may well break recursive commands sent with mon node ctrl.
if arg == '--':
rem_args += args[i:]
break
if arg.startswith('--merlin-cfg=') or arg.startswith('--config='):
mconf.config_file = arg.split('=', 1)[1]
elif arg.startswith('--type='):
wanted_types = arg.split('=')[1].split(',')
have_type_arg = True
elif arg.startswith('--name='):
wanted_names = arg.split('=')[1].split(',')
have_name_arg = True
else:
# not an argument we recognize, so stash it and move on
rem_args += [arg]
continue
# load the merlin configuration, and thus all nodes
mconf.parse()
return rem_args
def module_init(args):
global wanted_types, wanted_names
global have_type_arg, have_name_arg
global qh
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'query_socket':
qh = v[1]
break
wanted_types = mconf.merlin_node.valid_types
rem_args = []
i = -1
for arg in args:
i += 1
# stop parsing immediately if we hit double-dashes, or we
# may well break recursive commands sent with mon node ctrl.
if arg == '--':
rem_args += args[i:]
break
if arg.startswith('--merlin-cfg=') or arg.startswith('--config='):
mconf.config_file = arg.split('=', 1)[1]
elif arg.startswith('--type='):
wanted_types = arg.split('=')[1].split(',')
have_type_arg = True
elif arg.startswith('--name='):
wanted_names = arg.split('=')[1].split(',')
have_name_arg = True
else:
# not an argument we recognize, so stash it and move on
rem_args += [arg]
continue
# load the merlin configuration, and thus all nodes
mconf.parse()
return rem_args
def cmd_pasv(args):
"""
Submits passive checkresults to the nagios.cmd pipe and
verifies that the data gets written to database correctly
and in a timely manner.
Available options for 'mon test pasv'
--nagios-cfg=<file> default /opt/monitor/var/etc/nagios.cfg
--counters=<int> number of counters per object (default 30)
--hosts=<int> number of hosts (default 1)
--services=<int> number of services (default 5)
--loops=<int> number of loops (default 1)
--interval=<int> interval in seconds between loops (def 1800)
--delay=<int> delay between submitting and checking (def 25)
!!! WARNING !!! !!! WARNING !!!
This command will disble active checks on your system and have other
side-effects as well.
!!! WARNING !!! !!! WARNING !!!
"""
global verbose
nagios_cfg = False
num_hosts = 1
num_services = 5
num_loops = 1
num_counters = 30
interval = 1800
delay = 25
cmd_pipe = False
global send_host_checks
for arg in args:
if arg.startswith('--nagios-cfg='):
nagios_cfg = arg.split('=')[1]
elif arg.startswith('--counters='):
num_counters = int(arg.split('=')[1])
elif arg.startswith('--hosts='):
num_hosts = int(arg.split('=')[1])
elif arg.startswith('--services='):
num_services = int(arg.split('=')[1])
elif arg.startswith('--loops='):
num_loops = int(arg.split('=')[1])
elif arg.startswith('--interval='):
interval = int(arg.split('=')[1])
elif arg.startswith('--delay='):
delay = int(arg.split('=')[1])
elif arg == '--verbose' or arg == '-v':
verbose = True
elif arg == '--nohostchecks':
send_host_checks = False
else:
prettyprint_docstring("pasv", cmd_pasv.__doc__, "Unknown argument: %s" % arg)
if arg == '--help' or arg == 'help':
sys.exit(0)
else:
sys.exit(1)
if nagios_cfg:
comp = cconf.parse_conf(nagios_cfg)
for v in comp.params:
if v[0] == 'command_file':
cmd_pipe = v[1]
break
db = merlin_db.connect(mconf)
dbc = db.cursor()
if not cmd_pipe:
cmd_pipe = "/opt/monitor/var/rw/nagios.cmd"
cmd_fd = _pasv_open_cmdpipe(cmd_pipe)
# disable active checks while we test the passive ones, or
# active checkresults might overwrite the passive ones and
# contaminate the testresults
test_cmd(cmd_fd, "STOP_EXECUTING_HOST_CHECKS")
test_cmd(cmd_fd, "STOP_EXECUTING_SVC_CHECKS")
test_cmd(cmd_fd, "START_ACCEPTING_PASSIVE_HOST_CHECKS")
test_cmd(cmd_fd, "START_ACCEPTING_PASSIVE_SVC_CHECKS")
os.close(cmd_fd)
# now we update the database with impossible values so we
# know we don't get the right test-data by mistake in case
# the test-case is run multiple times directly following
# each other
dbc.execute("UPDATE host SET last_check = 5, current_state = 5")
dbc.execute("UPDATE service SET last_check = 5, current_state = 5")
host_list = []
test_objs = []
query = "SELECT host_name FROM host ORDER BY host_name ASC"
dbc.execute(query)
hi = 0
# arbitrary (very) large value
min_services = 100123098
min_services_host = ''
for row in dbc.fetchall():
if hi < num_hosts:
obj = pasv_test_host(row[0])
host_list.append(obj)
if send_host_checks:
test_objs.append(obj)
hi += 1
for host in host_list:
query = ("SELECT service_description FROM service "
"WHERE host_name = '%s' ORDER BY service_description ASC"
% host.name)
dbc.execute(query)
services = 0
si = 0
for row in dbc.fetchall():
if si < num_services:
services += 1
obj = pasv_test_service(host.name, row[0])
host.services.append(obj)
test_objs.append(obj)
si += 1
if services < min_services:
min_services_host = host.name
min_services = services
if num_hosts > host_list:
print("Can't run tests for %d hosts when only %d are configured" % (num_hosts, len(host_list)))
if num_services > min_services:
print("Can't run tests for %d services / host when %s has only %d configured"
% (num_services, min_services_host, min_services))
# primary testing loop
loops = 0
while loops < num_loops:
loops += 1
# generate the counters we'll be using.
# We get fresh ones for each iteration
counters = _generate_counters(num_counters)
cnt_string = "%s" % " ".join(counters)
cnt_hash = hashlib.sha(cnt_string).hexdigest()
# why we have to disconnect from db and re-open the
# command pipe is beyond me, but that's the way it is, it
# seems. It also matches real-world use a lot better,
# since the reader imitates ninja and the writer imitates
# nsca.
cmd_fd = _pasv_open_cmdpipe(cmd_pipe)
merlin_db.disconnect()
# new status every time so we can differ between the values
# and also test the worst-case scenario where the daemon has
# to run two queries for each passive checkresult
status = loops % 3
loop_start = time.time()
print("Submitting passive check results (%s) @ %s" % (cnt_hash, time.time()))
for t in test_objs:
cmd = _pasv_build_cmd(t, status)
cmd += "%s|%s\n" % (cnt_hash, cnt_string)
t.cmd_hash = hashlib.sha(cmd).hexdigest()
t.submit_time = time.time()
result = os.write(cmd_fd, cmd)
test(result, len(cmd), "%d of %d bytes written for %s" % (result, len(cmd), t.name))
os.close(cmd_fd)
db = merlin_db.connect(mconf)
dbc = db.cursor()
print("Sleeping %d seconds before reaping results" % delay)
time.sleep(delay)
for t in test_objs:
query = ("SELECT "
"last_check, current_state, output, perf_data "
"FROM %s" % t.query())
dbc.execute(query)
row = dbc.fetchone()
test(row[0] + delay > t.submit_time, True, "reasonable delay for %s" % t.name)
test(row[1], status, "status updated for %s" % t.name)
test(str(row[2]), cnt_hash, "output update for %s" % t.name)
test(str(row[3]), cnt_string, "counter truncation check for %s" % t.name)
if loops < num_loops:
interval_sleep = (loop_start + interval) - time.time()
if interval_sleep > 0:
print("Sleeping %d seconds until next test-set" % interval_sleep)
time.sleep(interval_sleep)
total_tests = failed + passed
print("failed: %d/%.3f%%" % (failed, float(failed * 100) / total_tests))
print("passed: %d/%.3f%%" % (passed, float(passed * 100) / total_tests))
print("total tests: %d" % total_tests)
def _import_objects_from_cache(self, inputfile, is_cache=True):
'''Given a file name, insert all objects in it into the database'''
last_obj_type = ''
obj_array = {}
if not self.tables_truncated:
for table in self.tables_to_truncate:
self.db.sql_exec('TRUNCATE ' + table)
self.tables_truncated = True
service_slaves = {'serviceescalation': 1, 'servicedependency': 1}
try:
root = parse_conf(inputfile, (is_cache and '\t' or '='))
except IOError:
print "Error: Couldn't open file '%s'." % inputfile
return
if not root.objects:
print "Couldn't parse file."
return
if root.objects[0].name in ('info', 'program'):
self.importing_status = True
for obj in root.objects:
obj_type = obj.name
if obj_type.startswith('define '):
obj_type = obj_type[7:]
if self.conv_type.has_key(obj_type):
obj_type = self.conv_type[obj_type]
if obj_type != last_obj_type:
obj_array = self._done_parsing_obj_type_objects(last_obj_type, obj_array)
last_obj_type = obj_type
obj_data = self._mangle_data(obj_type, obj.params)
if not obj_data:
continue
obj_name = self._get_obj_name(obj_type, obj_data)
old_obj = self.indexer.get(obj_type, obj_name)
if not old_obj:
old_obj = Entry()
old_obj.name = obj_name
if not self.base_oid.has_key(obj_type):
self.base_oid[obj_type] = 1
old_obj.id = self.base_oid[obj_type]
self.base_oid[obj_type] += 1
obj_data['is a fresh one'] = True
if obj_name:
self.indexer.set(obj_type, obj_name, old_obj.id)
if old_obj.instance_id != 0:
obj_data['instance_id'] = old_obj.instance_id
obj_key = old_obj.id
if obj_type in ('host', 'program_status'):
pass
elif obj_type == 'timeperiod':
pass
elif obj_type not in ('hostgroup', 'servicegroup', 'contactgroup'):
if service_slaves.has_key(obj_type):
obj_data = self._post_mangle_service_slave(obj_type,
obj_data)
if obj_data:
if not obj_array.has_key(obj_type):
obj_array[obj_type] = {}
obj_array[obj_type][obj_key] = obj_data
self.obj_array = self._done_parsing_obj_type_objects(obj_type, obj_array)
def _import_objects_from_cache(self, inputfile, is_cache=True):
'''Given a file name, insert all objects in it into the database'''
last_obj_type = ''
obj_array = {}
if not self.tables_truncated:
for table in self.tables_to_truncate:
self.db.sql_exec('TRUNCATE ' + table)
self.tables_truncated = True
service_slaves = {'serviceescalation': 1, 'servicedependency': 1}
try:
root = parse_conf(inputfile, (is_cache and '\t' or '='))
except IOError:
print "Error: Couldn't open file '%s'." % inputfile
return
if not root.objects:
print "Couldn't parse file."
return
if root.objects[0].name in ('info', 'program'):
self.importing_status = True
for obj in root.objects:
obj_type = obj.name
if obj_type.startswith('define '):
obj_type = obj_type[7:]
if self.conv_type.has_key(obj_type):
obj_type = self.conv_type[obj_type]
if obj_type != last_obj_type:
obj_array = self._done_parsing_obj_type_objects(
last_obj_type, obj_array)
last_obj_type = obj_type
obj_data = self._mangle_data(obj_type, obj.params)
if not obj_data:
continue
obj_name = self._get_obj_name(obj_type, obj_data)
old_obj = self.indexer.get(obj_type, obj_name)
if not old_obj:
old_obj = Entry()
old_obj.name = obj_name
if not self.base_oid.has_key(obj_type):
self.base_oid[obj_type] = 1
old_obj.id = self.base_oid[obj_type]
self.base_oid[obj_type] += 1
obj_data['is a fresh one'] = True
if obj_name:
self.indexer.set(obj_type, obj_name, old_obj.id)
if old_obj.instance_id != 0:
obj_data['instance_id'] = old_obj.instance_id
obj_key = old_obj.id
if obj_type in ('host', 'program_status'):
pass
elif obj_type == 'timeperiod':
pass
elif obj_type not in ('hostgroup', 'servicegroup', 'contactgroup'):
if service_slaves.has_key(obj_type):
obj_data = self._post_mangle_service_slave(
obj_type, obj_data)
if obj_data:
if not obj_array.has_key(obj_type):
obj_array[obj_type] = {}
obj_array[obj_type][obj_key] = obj_data
self.obj_array = self._done_parsing_obj_type_objects(
obj_type, obj_array)
