def test_discover_set_name_from_primary(self):
# Discovering a replica set without the setName supplied by the user
# is not yet supported by MongoClient, but Topology can do it.
topology_settings = SetNameDiscoverySettings(
seeds=[address],
pool_class=MockPool,
monitor_class=MockMonitor)
t = Topology(topology_settings)
self.assertEqual(t.description.replica_set_name, None)
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
t.open()
got_ismaster(t, address, {
'ok': 1,
'ismaster': True,
'setName': 'rs',
'hosts': ['a']})
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetWithPrimary)
# Another response from the primary. Tests the code that processes
# primary response when topology type is already ReplicaSetWithPrimary.
got_ismaster(t, address, {
'ok': 1,
'ismaster': True,
'setName': 'rs',
'hosts': ['a']})
# No change.
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetWithPrimary)
def create_mock_topology(
seeds=None,
replica_set_name=None,
monitor_class=MockMonitor):
partitioned_seeds = list(imap(common.partition_node, seeds or ['a']))
topology_settings = TopologySettings(
partitioned_seeds,
replica_set_name=replica_set_name,
pool_class=MockPool,
monitor_class=monitor_class)
t = Topology(topology_settings)
t.open()
return t
def run_scenario(self):
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
# Initialize topologies.
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
topology = Topology(
TopologySettings(seeds=seeds,
monitor_class=MockMonitor,
pool_class=MockPool,
heartbeat_frequency=frequency))
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Create server selector.
# Make first letter lowercase to match read_pref's modes.
pref_def = scenario_def['read_preference']
mode_string = pref_def.get('mode', 'primary')
mode_string = mode_string[:1].lower() + mode_string[1:]
mode = read_preferences.read_pref_mode_from_name(mode_string)
max_staleness = pref_def.get('maxStalenessMS', 0) / 1000.0
tag_sets = pref_def.get('tag_sets')
if scenario_def.get('error'):
with self.assertRaises(ConfigurationError):
# Error can be raised when making Read Pref or selecting.
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
topology.select_server(pref)
return
expected_addrs = set([
server['address'] for server in scenario_def['in_latency_window']])
# Select servers.
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
if not expected_addrs:
with self.assertRaises(ConnectionFailure):
topology.select_servers(pref, server_selection_timeout=0)
return
servers = topology.select_servers(pref, server_selection_timeout=0)
actual_addrs = set(['%s:%d' % s.description.address for s in servers])
for unexpected in actual_addrs - expected_addrs:
self.fail("'%s' shouldn't have been selected, but was" % unexpected)
for unselected in expected_addrs - actual_addrs:
self.fail("'%s' should have been selected, but wasn't" % unselected)
def create_mock_topology(uri, monitor_class=MockMonitor):
# Some tests in the spec include URIs like mongodb://A/?connect=direct,
# but PyMongo considers any single-seed URI with no setName to be "direct".
parsed_uri = parse_uri(uri.replace('connect=direct', ''))
replica_set_name = None
if 'replicaset' in parsed_uri['options']:
replica_set_name = parsed_uri['options']['replicaset']
topology_settings = TopologySettings(
parsed_uri['nodelist'],
replica_set_name=replica_set_name,
pool_class=MockPool,
monitor_class=monitor_class)
c = Topology(topology_settings)
c.open()
return c
def test_timeout_configuration(self):
pool_options = PoolOptions(connect_timeout=1, socket_timeout=2)
topology_settings = TopologySettings(pool_options=pool_options)
t = Topology(topology_settings=topology_settings)
t.open()
# Get the default server.
server = t.get_server_by_address(('localhost', 27017))
# The pool for application operations obeys our settings.
self.assertEqual(1, server._pool.opts.connect_timeout)
self.assertEqual(2, server._pool.opts.socket_timeout)
# The pool for monitoring operations uses our connect_timeout as both
# its connect_timeout and its socket_timeout.
monitor = server._monitor
self.assertEqual(1, monitor._pool.opts.connect_timeout)
self.assertEqual(1, monitor._pool.opts.socket_timeout)
# The monitor, not its pool, is responsible for calling ismaster.
self.assertFalse(monitor._pool.handshake)
def test_timeout_configuration(self):
pool_options = PoolOptions(connect_timeout=1, socket_timeout=2)
topology_settings = TopologySettings(pool_options=pool_options)
t = Topology(topology_settings=topology_settings)
t.open()
# Get the default server.
server = t.get_server_by_address(('localhost', 27017))
# The pool for application operations obeys our settings.
self.assertEqual(1, server._pool.opts.connect_timeout)
self.assertEqual(2, server._pool.opts.socket_timeout)
# The pool for monitoring operations uses our connect_timeout as both
# its connect_timeout and its socket_timeout.
monitor = server._monitor
self.assertEqual(1, monitor._pool.opts.connect_timeout)
self.assertEqual(1, monitor._pool.opts.socket_timeout)
# The monitor, not its pool, is responsible for calling ismaster.
self.assertFalse(monitor._pool.handshake)
def create_topology(scenario_def, **kwargs):
# Initialize topologies.
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
settings = get_topology_settings_dict(heartbeat_frequency=frequency,
seeds=seeds,
**kwargs)
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
topology = Topology(TopologySettings(**settings))
topology.open()
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
return topology
def test_server_selector_bypassed(self):
selector = FunctionCallCounter(lambda x: x)
scenario_def = {
'topology_description': {
'type':
'ReplicaSetNoPrimary',
'servers': [
{
'address': 'b:27017',
'avg_rtt_ms': 10000,
'type': 'RSSecondary',
'tag': {}
},
{
'address': 'c:27017',
'avg_rtt_ms': 20000,
'type': 'RSSecondary',
'tag': {}
},
{
'address': 'a:27017',
'avg_rtt_ms': 30000,
'type': 'RSSecondary',
'tag': {}
},
]
}
}
# Create & populate Topology such that no server is writeable.
seeds, hosts = get_addresses(
scenario_def["topology_description"]["servers"])
settings = get_topology_settings_dict(heartbeat_frequency=1,
local_threshold_ms=1,
seeds=seeds,
server_selector=selector)
topology = Topology(TopologySettings(**settings))
topology.open()
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Invoke server selection and assert no calls to our custom selector.
with self.assertRaisesRegex(ServerSelectionTimeoutError,
'No primary available for writes'):
topology.select_server(writable_server_selector,
server_selection_timeout=0.1)
self.assertEqual(selector.call_count, 0)
def create_mock_topology(uri, monitor_class=DummyMonitor):
parsed_uri = parse_uri(uri)
replica_set_name = None
direct_connection = None
load_balanced = None
if 'replicaset' in parsed_uri['options']:
replica_set_name = parsed_uri['options']['replicaset']
if 'directConnection' in parsed_uri['options']:
direct_connection = parsed_uri['options']['directConnection']
if 'loadBalanced' in parsed_uri['options']:
load_balanced = parsed_uri['options']['loadBalanced']
topology_settings = TopologySettings(
parsed_uri['nodelist'],
replica_set_name=replica_set_name,
monitor_class=monitor_class,
direct_connection=direct_connection,
load_balanced=load_balanced)
c = Topology(topology_settings)
c.open()
return c
def test_discover_set_name_from_secondary(self):
# Discovering a replica set without the setName supplied by the user
# is not yet supported by MongoClient, but Topology can do it.
topology_settings = SetNameDiscoverySettings(
seeds=[address],
pool_class=MockPool,
monitor_class=MockMonitor)
t = Topology(topology_settings)
self.assertEqual(t.description.replica_set_name, None)
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
t.open()
got_ismaster(t, address, {
'ok': 1,
'ismaster': False,
'secondary': True,
'setName': 'rs',
'hosts': ['a']})
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
def test_discover_set_name_from_secondary(self):
# Discovering a replica set without the setName supplied by the user
# is not yet supported by MongoClient, but Topology can do it.
topology_settings = SetNameDiscoverySettings(
seeds=[address], pool_class=MockPool, monitor_class=DummyMonitor)
t = Topology(topology_settings)
self.assertEqual(t.description.replica_set_name, None)
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
t.open()
got_ismaster(
t, address, {
'ok': 1,
'ismaster': False,
'secondary': True,
'setName': 'rs',
'hosts': ['a']
})
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
def test_discover_set_name_from_primary(self):
# Discovering a replica set without the setName supplied by the user
# is not yet supported by MongoClient, but Topology can do it.
topology_settings = SetNameDiscoverySettings(seeds=[address],
pool_class=MockPool,
monitor_class=MockMonitor)
t = Topology(topology_settings)
self.assertEqual(t.description.replica_set_name, None)
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetNoPrimary)
t.open()
got_ismaster(t, address, {
'ok': 1,
'ismaster': True,
'setName': 'rs',
'hosts': ['a']
})
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetWithPrimary)
# Another response from the primary. Tests the code that processes
# primary response when topology type is already ReplicaSetWithPrimary.
got_ismaster(t, address, {
'ok': 1,
'ismaster': True,
'setName': 'rs',
'hosts': ['a']
})
# No change.
self.assertEqual(t.description.replica_set_name, 'rs')
self.assertEqual(t.description.topology_type,
TOPOLOGY_TYPE.ReplicaSetWithPrimary)
def test_latency_threshold_application(self):
selector = SelectionStoreSelector()
scenario_def = {
'topology_description': {
'type': 'ReplicaSetWithPrimary', 'servers': [
{'address': 'b:27017',
'avg_rtt_ms': 10000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'c:27017',
'avg_rtt_ms': 20000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'a:27017',
'avg_rtt_ms': 30000,
'type': 'RSPrimary',
'tag': {}},
]}}
# Create & populate Topology such that all but one server is too slow.
rtt_times = [srv['avg_rtt_ms'] for srv in
scenario_def['topology_description']['servers']]
min_rtt_idx = rtt_times.index(min(rtt_times))
seeds, hosts = get_addresses(
scenario_def["topology_description"]["servers"])
settings = get_topology_settings_dict(
heartbeat_frequency=1, local_threshold_ms=1, seeds=seeds,
server_selector=selector)
topology = Topology(TopologySettings(**settings))
topology.open()
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Invoke server selection and assert no filtering based on latency
# prior to custom server selection logic kicking in.
server = topology.select_server(ReadPreference.NEAREST)
self.assertEqual(
len(selector.selection),
len(topology.description.server_descriptions()))
# Ensure proper filtering based on latency after custom selection.
self.assertEqual(
server.description.address, seeds[min_rtt_idx])
def test_latency_threshold_application(self):
selector = SelectionStoreSelector()
scenario_def = {
'topology_description': {
'type': 'ReplicaSetWithPrimary', 'servers': [
{'address': 'b:27017',
'avg_rtt_ms': 10000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'c:27017',
'avg_rtt_ms': 20000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'a:27017',
'avg_rtt_ms': 30000,
'type': 'RSPrimary',
'tag': {}},
]}}
# Create & populate Topology such that all but one server is too slow.
rtt_times = [srv['avg_rtt_ms'] for srv in
scenario_def['topology_description']['servers']]
min_rtt_idx = rtt_times.index(min(rtt_times))
seeds, hosts = get_addresses(
scenario_def["topology_description"]["servers"])
settings = get_topology_settings_dict(
heartbeat_frequency=1, local_threshold_ms=1, seeds=seeds,
server_selector=selector)
topology = Topology(TopologySettings(**settings))
topology.open()
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Invoke server selection and assert no filtering based on latency
# prior to custom server selection logic kicking in.
server = topology.select_server(ReadPreference.NEAREST)
self.assertEqual(
len(selector.selection),
len(topology.description.server_descriptions()))
# Ensure proper filtering based on latency after custom selection.
self.assertEqual(
server.description.address, seeds[min_rtt_idx])
def test_server_selector_bypassed(self):
selector = CallCountSelector()
scenario_def = {
'topology_description': {
'type': 'ReplicaSetNoPrimary', 'servers': [
{'address': 'b:27017',
'avg_rtt_ms': 10000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'c:27017',
'avg_rtt_ms': 20000,
'type': 'RSSecondary',
'tag': {}},
{'address': 'a:27017',
'avg_rtt_ms': 30000,
'type': 'RSSecondary',
'tag': {}},
]}}
# Create & populate Topology such that no server is writeable.
seeds, hosts = get_addresses(
scenario_def["topology_description"]["servers"])
settings = get_topology_settings_dict(
heartbeat_frequency=1, local_threshold_ms=1, seeds=seeds,
server_selector=selector)
topology = Topology(TopologySettings(**settings))
topology.open()
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Invoke server selection and assert no calls to our custom selector.
with self.assertRaisesRegex(
ServerSelectionTimeoutError, 'No primary available for writes'):
topology.select_server(
writable_server_selector, server_selection_timeout=0.1)
self.assertEqual(selector.call_count, 0)
def create_topology(scenario_def, **kwargs):
# Initialize topologies.
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
topology_type = get_topology_type_name(scenario_def)
if topology_type == 'LoadBalanced':
kwargs.setdefault('load_balanced', True)
# Force topology description to ReplicaSet
elif topology_type in ['ReplicaSetNoPrimary', 'ReplicaSetWithPrimary']:
kwargs.setdefault('replica_set_name', 'rs')
settings = get_topology_settings_dict(heartbeat_frequency=frequency,
seeds=seeds,
**kwargs)
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
topology = Topology(TopologySettings(**settings))
topology.open()
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Assert that descriptions match
assert (
scenario_def['topology_description']['type'] == topology.description.
topology_type_name), topology.description.topology_type_name
return topology
def run_scenario(self):
# Initialize topologies.
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
settings = dict(monitor_class=MockMonitor,
heartbeat_frequency=frequency,
pool_class=MockPool)
settings['seeds'], hosts = get_addresses(
scenario_def['topology_description']['servers'])
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
top_latency = Topology(TopologySettings(**settings))
top_latency.open()
# "In latency window" is defined in the server selection
# spec as the subset of suitable_servers that falls within the
# allowable latency window.
settings['local_threshold_ms'] = 1000000
top_suitable = Topology(TopologySettings(**settings))
top_suitable.open()
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
top_suitable.on_change(server_description)
top_latency.on_change(server_description)
# Create server selector.
if scenario_def.get("operation") == "write":
pref = writable_server_selector
else:
# Make first letter lowercase to match read_pref's modes.
pref_def = scenario_def['read_preference']
mode_string = pref_def.get('mode', 'primary')
mode_string = mode_string[:1].lower() + mode_string[1:]
mode = read_preferences.read_pref_mode_from_name(mode_string)
max_staleness = pref_def.get('maxStalenessSeconds', -1)
tag_sets = pref_def.get('tag_sets')
if scenario_def.get('error'):
with self.assertRaises((ConfigurationError, ValueError)):
# Error can be raised when making Read Pref or selecting.
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
top_latency.select_server(pref)
return
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
# Select servers.
if not scenario_def.get('suitable_servers'):
with self.assertRaises(AutoReconnect):
top_suitable.select_server(pref, server_selection_timeout=0)
return
if not scenario_def['in_latency_window']:
with self.assertRaises(AutoReconnect):
top_latency.select_server(pref, server_selection_timeout=0)
return
actual_suitable_s = top_suitable.select_servers(
pref, server_selection_timeout=0)
actual_latency_s = top_latency.select_servers(
pref, server_selection_timeout=0)
expected_suitable_servers = {}
for server in scenario_def['suitable_servers']:
server_description = make_server_description(server, hosts)
expected_suitable_servers[server['address']] = server_description
actual_suitable_servers = {}
for s in actual_suitable_s:
actual_suitable_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_suitable_servers),
len(expected_suitable_servers))
for k, actual in actual_suitable_servers.items():
expected = expected_suitable_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
expected_latency_servers = {}
for server in scenario_def['in_latency_window']:
server_description = make_server_description(server, hosts)
expected_latency_servers[server['address']] = server_description
actual_latency_servers = {}
for s in actual_latency_s:
actual_latency_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_latency_servers),
len(expected_latency_servers))
for k, actual in actual_latency_servers.items():
expected = expected_latency_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
def run_scenario(self):
# Initialize topologies.
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
settings = get_topology_settings_dict(
heartbeat_frequency=frequency,
seeds=seeds
)
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
top_latency = Topology(TopologySettings(**settings))
top_latency.open()
# "In latency window" is defined in the server selection
# spec as the subset of suitable_servers that falls within the
# allowable latency window.
settings['local_threshold_ms'] = 1000000
top_suitable = Topology(TopologySettings(**settings))
top_suitable.open()
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
top_suitable.on_change(server_description)
top_latency.on_change(server_description)
# Create server selector.
if scenario_def.get("operation") == "write":
pref = writable_server_selector
else:
# Make first letter lowercase to match read_pref's modes.
pref_def = scenario_def['read_preference']
if scenario_def.get('error'):
with self.assertRaises((ConfigurationError, ValueError)):
# Error can be raised when making Read Pref or selecting.
pref = parse_read_preference(pref_def)
top_latency.select_server(pref)
return
pref = parse_read_preference(pref_def)
# Select servers.
if not scenario_def.get('suitable_servers'):
with self.assertRaises(AutoReconnect):
top_suitable.select_server(pref, server_selection_timeout=0)
return
if not scenario_def['in_latency_window']:
with self.assertRaises(AutoReconnect):
top_latency.select_server(pref, server_selection_timeout=0)
return
actual_suitable_s = top_suitable.select_servers(
pref, server_selection_timeout=0)
actual_latency_s = top_latency.select_servers(
pref, server_selection_timeout=0)
expected_suitable_servers = {}
for server in scenario_def['suitable_servers']:
server_description = make_server_description(server, hosts)
expected_suitable_servers[server['address']] = server_description
actual_suitable_servers = {}
for s in actual_suitable_s:
actual_suitable_servers["%s:%d" % (s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_suitable_servers),
len(expected_suitable_servers))
for k, actual in actual_suitable_servers.items():
expected = expected_suitable_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
expected_latency_servers = {}
for server in scenario_def['in_latency_window']:
server_description = make_server_description(server, hosts)
expected_latency_servers[server['address']] = server_description
actual_latency_servers = {}
for s in actual_latency_s:
actual_latency_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_latency_servers),
len(expected_latency_servers))
for k, actual in actual_latency_servers.items():
expected = expected_latency_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
def __init__(self,
host=None,
port=None,
document_class=dict,
tz_aware=False,
connect=True,
**kwargs):
"""Client for a MongoDB instance, a replica set, or a set of mongoses.
The client object is thread-safe and has connection-pooling built in.
If an operation fails because of a network error,
:class:`~pymongo.errors.ConnectionFailure` is raised and the client
reconnects in the background. Application code should handle this
exception (recognizing that the operation failed) and then continue to
execute.
The `host` parameter can be a full `mongodb URI
<http://dochub.mongodb.org/core/connections>`_, in addition to
a simple hostname. It can also be a list of hostnames or
URIs. Any port specified in the host string(s) will override
the `port` parameter. If multiple mongodb URIs containing
database or auth information are passed, the last database,
username, and password present will be used. For username and
passwords reserved characters like ':', '/', '+' and '@' must be
escaped following RFC 2396.
.. warning:: When using PyMongo in a multiprocessing context, please
read :ref:`multiprocessing` first.
:Parameters:
- `host` (optional): hostname or IP address of the
instance to connect to, or a mongodb URI, or a list of
hostnames / mongodb URIs. If `host` is an IPv6 literal
it must be enclosed in '[' and ']' characters following
the RFC2732 URL syntax (e.g. '[::1]' for localhost)
- `port` (optional): port number on which to connect
- `document_class` (optional): default class to use for
documents returned from queries on this client
- `tz_aware` (optional): if ``True``,
:class:`~datetime.datetime` instances returned as values
in a document by this :class:`MongoClient` will be timezone
aware (otherwise they will be naive)
- `connect` (optional): if ``True`` (the default), immediately
begin connecting to MongoDB in the background. Otherwise connect
on the first operation.
| **Other optional parameters can be passed as keyword arguments:**
- `maxPoolSize` (optional): The maximum number of connections
that the pool will open simultaneously. If this is set, operations
will block if there are `maxPoolSize` outstanding connections
from the pool. Defaults to 100. Cannot be 0.
- `socketTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait for a response after sending an
ordinary (non-monitoring) database operation before concluding that
a network error has occurred. Defaults to ``None`` (no timeout).
- `connectTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait during server monitoring when
connecting a new socket to a server before concluding the server
is unavailable. Defaults to ``20000`` (20 seconds).
- `serverSelectionTimeoutMS`: (integer) Controls how long (in
milliseconds) the driver will wait to find an available,
appropriate server to carry out a database operation; while it is
waiting, multiple server monitoring operations may be carried out,
each controlled by `connectTimeoutMS`. Defaults to ``30000`` (30
seconds).
- `waitQueueTimeoutMS`: (integer or None) How long (in milliseconds)
a thread will wait for a socket from the pool if the pool has no
free sockets. Defaults to ``None`` (no timeout).
- `waitQueueMultiple`: (integer or None) Multiplied by maxPoolSize
to give the number of threads allowed to wait for a socket at one
time. Defaults to ``None`` (no limit).
- `socketKeepAlive`: (boolean) Whether to send periodic keep-alive
packets on connected sockets. Defaults to ``False`` (do not send
keep-alive packets).
| **Write Concern options:**
| (Only set if passed. No default values.)
- `w`: (integer or string) If this is a replica set, write operations
will block until they have been replicated to the specified number
or tagged set of servers. `w=<int>` always includes the replica set
primary (e.g. w=3 means write to the primary and wait until
replicated to **two** secondaries). Passing w=0 **disables write
acknowledgement** and all other write concern options.
- `wtimeout`: (integer) Used in conjunction with `w`. Specify a value
in milliseconds to control how long to wait for write propagation
to complete. If replication does not complete in the given
timeframe, a timeout exception is raised.
- `j`: If ``True`` block until write operations have been committed
to the journal. Cannot be used in combination with `fsync`. Prior
to MongoDB 2.6 this option was ignored if the server was running
without journaling. Starting with MongoDB 2.6 write operations will
fail with an exception if this option is used when the server is
running without journaling.
- `fsync`: If ``True`` and the server is running without journaling,
blocks until the server has synced all data files to disk. If the
server is running with journaling, this acts the same as the `j`
option, blocking until write operations have been committed to the
journal. Cannot be used in combination with `j`.
| **Replica set keyword arguments for connecting with a replica set
- either directly or via a mongos:**
- `replicaSet`: (string or None) The name of the replica set to
connect to. The driver will verify that all servers it connects to
match this name. Implies that the hosts specified are a seed list
and the driver should attempt to find all members of the set.
Defaults to ``None``.
- `read_preference`: The read preference for this client. If
connecting directly to a secondary then a read preference mode
*other* than PRIMARY is required - otherwise all queries will throw
:class:`~pymongo.errors.AutoReconnect` "not master".
See :class:`~pymongo.read_preferences.ReadPreference` for all
available read preference options. Defaults to ``PRIMARY``.
| **SSL configuration:**
- `ssl`: If ``True``, create the connection to the server using SSL.
Defaults to ``False``.
- `ssl_keyfile`: The private keyfile used to identify the local
connection against mongod. If included with the ``certfile`` then
only the ``ssl_certfile`` is needed. Implies ``ssl=True``.
Defaults to ``None``.
- `ssl_certfile`: The certificate file used to identify the local
connection against mongod. Implies ``ssl=True``. Defaults to
``None``.
- `ssl_cert_reqs`: Specifies whether a certificate is required from
the other side of the connection, and whether it will be validated
if provided. It must be one of the three values ``ssl.CERT_NONE``
(certificates ignored), ``ssl.CERT_OPTIONAL``
(not required, but validated if provided), or ``ssl.CERT_REQUIRED``
(required and validated). If the value of this parameter is not
``ssl.CERT_NONE`` and a value is not provided for ``ssl_ca_certs``
PyMongo will attempt to load system provided CA certificates.
If the python version in use does not support loading system CA
certificates then the ``ssl_ca_certs`` parameter must point
to a file of CA certificates. Implies ``ssl=True``. Defaults to
``ssl.CERT_REQUIRED`` if not provided and ``ssl=True``.
- `ssl_ca_certs`: The ca_certs file contains a set of concatenated
"certification authority" certificates, which are used to validate
certificates passed from the other end of the connection.
Implies ``ssl=True``. Defaults to ``None``.
- `ssl_match_hostname`: If ``True`` (the default), and
`ssl_cert_reqs` is not ``ssl.CERT_NONE``, enables hostname
verification using the :func:`~ssl.match_hostname` function from
python's :mod:`~ssl` module. Think very carefully before setting
this to ``False`` as that could make your application vulnerable to
man-in-the-middle attacks.
.. mongodoc:: connections
.. versionchanged:: 3.0
:class:`~pymongo.mongo_client.MongoClient` is now the one and only
client class for a standalone server, mongos, or replica set.
It includes the functionality that had been split into
:class:`~pymongo.mongo_client.MongoReplicaSetClient`: it can connect
to a replica set, discover all its members, and monitor the set for
stepdowns, elections, and reconfigs.
The :class:`~pymongo.mongo_client.MongoClient` constructor no
longer blocks while connecting to the server or servers, and it no
longer raises :class:`~pymongo.errors.ConnectionFailure` if they
are unavailable, nor :class:`~pymongo.errors.ConfigurationError`
if the user's credentials are wrong. Instead, the constructor
returns immediately and launches the connection process on
background threads.
Therefore the ``alive`` method is removed since it no longer
provides meaningful information; even if the client is disconnected,
it may discover a server in time to fulfill the next operation.
In PyMongo 2.x, :class:`~pymongo.MongoClient` accepted a list of
standalone MongoDB servers and used the first it could connect to::
MongoClient(['host1.com:27017', 'host2.com:27017'])
A list of multiple standalones is no longer supported; if multiple
servers are listed they must be members of the same replica set, or
mongoses in the same sharded cluster.
The behavior for a list of mongoses is changed from "high
availability" to "load balancing". Before, the client connected to
the lowest-latency mongos in the list, and used it until a network
error prompted it to re-evaluate all mongoses' latencies and
reconnect to one of them. In PyMongo 3, the client monitors its
network latency to all the mongoses continuously, and distributes
operations evenly among those with the lowest latency. See
:ref:`mongos-load-balancing` for more information.
The ``connect`` option is added.
The ``start_request``, ``in_request``, and ``end_request`` methods
are removed, as well as the ``auto_start_request`` option.
The ``copy_database`` method is removed, see the
:doc:`copy_database examples </examples/copydb>` for alternatives.
The :meth:`MongoClient.disconnect` method is removed; it was a
synonym for :meth:`~pymongo.MongoClient.close`.
:class:`~pymongo.mongo_client.MongoClient` no longer returns an
instance of :class:`~pymongo.database.Database` for attribute names
with leading underscores. You must use dict-style lookups instead::
client['__my_database__']
Not::
client.__my_database__
"""
if host is None:
host = self.HOST
if isinstance(host, string_type):
host = [host]
if port is None:
port = self.PORT
if not isinstance(port, int):
raise TypeError("port must be an instance of int")
seeds = set()
username = None
password = None
dbase = None
opts = {}
for entity in host:
if "://" in entity:
if entity.startswith("mongodb://"):
res = uri_parser.parse_uri(entity, port, warn=True)
seeds.update(res["nodelist"])
username = res["username"] or username
password = res["password"] or password
dbase = res["database"] or dbase
opts = res["options"]
else:
idx = entity.find("://")
raise InvalidURI("Invalid URI scheme: "
"%s" % (entity[:idx], ))
else:
seeds.update(uri_parser.split_hosts(entity, port))
if not seeds:
raise ConfigurationError("need to specify at least one host")
# _pool_class, _monitor_class, and _condition_class are for deep
# customization of PyMongo, e.g. Motor.
pool_class = kwargs.pop('_pool_class', None)
monitor_class = kwargs.pop('_monitor_class', None)
condition_class = kwargs.pop('_condition_class', None)
keyword_opts = kwargs
keyword_opts['document_class'] = document_class
keyword_opts['tz_aware'] = tz_aware
keyword_opts['connect'] = connect
# Validate all keyword options.
keyword_opts = dict(
common.validate(k, v) for k, v in keyword_opts.items())
opts.update(keyword_opts)
self.__options = options = ClientOptions(username, password, dbase,
opts)
self.__default_database_name = dbase
self.__lock = threading.Lock()
self.__cursor_manager = CursorManager(self)
self.__kill_cursors_queue = []
# Cache of existing indexes used by ensure_index ops.
self.__index_cache = {}
super(MongoClient,
self).__init__(options.codec_options, options.read_preference,
options.write_concern)
self.__all_credentials = {}
creds = options.credentials
if creds:
self._cache_credentials(creds.source, creds)
self._topology_settings = TopologySettings(
seeds=seeds,
replica_set_name=options.replica_set_name,
pool_class=pool_class,
pool_options=options.pool_options,
monitor_class=monitor_class,
condition_class=condition_class,
local_threshold_ms=options.local_threshold_ms,
server_selection_timeout=options.server_selection_timeout)
self._topology = Topology(self._topology_settings)
if connect:
self._topology.open()
def target():
client = self_ref()
if client is None:
return False # Stop the executor.
MongoClient._process_kill_cursors_queue(client)
return True
executor = periodic_executor.PeriodicExecutor(
condition_class=self._topology_settings.condition_class,
interval=common.KILL_CURSOR_FREQUENCY,
min_interval=0,
target=target,
name="pymongo_kill_cursors_thread")
# We strongly reference the executor and it weakly references us via
# this closure. When the client is freed, stop the executor soon.
self_ref = weakref.ref(self, executor.close)
self._kill_cursors_executor = executor
executor.open()
class MongoClient(common.BaseObject):
HOST = "localhost"
PORT = 27017
# Define order to retrieve options from ClientOptions for __repr__.
# No host/port; these are retrieved from TopologySettings.
_constructor_args = ('document_class', 'tz_aware', 'connect')
def __init__(self,
host=None,
port=None,
document_class=dict,
tz_aware=False,
connect=True,
**kwargs):
"""Client for a MongoDB instance, a replica set, or a set of mongoses.
The client object is thread-safe and has connection-pooling built in.
If an operation fails because of a network error,
:class:`~pymongo.errors.ConnectionFailure` is raised and the client
reconnects in the background. Application code should handle this
exception (recognizing that the operation failed) and then continue to
execute.
The `host` parameter can be a full `mongodb URI
<http://dochub.mongodb.org/core/connections>`_, in addition to
a simple hostname. It can also be a list of hostnames or
URIs. Any port specified in the host string(s) will override
the `port` parameter. If multiple mongodb URIs containing
database or auth information are passed, the last database,
username, and password present will be used. For username and
passwords reserved characters like ':', '/', '+' and '@' must be
escaped following RFC 2396.
.. warning:: When using PyMongo in a multiprocessing context, please
read :ref:`multiprocessing` first.
:Parameters:
- `host` (optional): hostname or IP address of the
instance to connect to, or a mongodb URI, or a list of
hostnames / mongodb URIs. If `host` is an IPv6 literal
it must be enclosed in '[' and ']' characters following
the RFC2732 URL syntax (e.g. '[::1]' for localhost)
- `port` (optional): port number on which to connect
- `document_class` (optional): default class to use for
documents returned from queries on this client
- `tz_aware` (optional): if ``True``,
:class:`~datetime.datetime` instances returned as values
in a document by this :class:`MongoClient` will be timezone
aware (otherwise they will be naive)
- `connect` (optional): if ``True`` (the default), immediately
begin connecting to MongoDB in the background. Otherwise connect
on the first operation.
| **Other optional parameters can be passed as keyword arguments:**
- `maxPoolSize` (optional): The maximum number of connections
that the pool will open simultaneously. If this is set, operations
will block if there are `maxPoolSize` outstanding connections
from the pool. Defaults to 100. Cannot be 0.
- `socketTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait for a response after sending an
ordinary (non-monitoring) database operation before concluding that
a network error has occurred. Defaults to ``None`` (no timeout).
- `connectTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait during server monitoring when
connecting a new socket to a server before concluding the server
is unavailable. Defaults to ``20000`` (20 seconds).
- `serverSelectionTimeoutMS`: (integer) Controls how long (in
milliseconds) the driver will wait to find an available,
appropriate server to carry out a database operation; while it is
waiting, multiple server monitoring operations may be carried out,
each controlled by `connectTimeoutMS`. Defaults to ``30000`` (30
seconds).
- `waitQueueTimeoutMS`: (integer or None) How long (in milliseconds)
a thread will wait for a socket from the pool if the pool has no
free sockets. Defaults to ``None`` (no timeout).
- `waitQueueMultiple`: (integer or None) Multiplied by maxPoolSize
to give the number of threads allowed to wait for a socket at one
time. Defaults to ``None`` (no limit).
- `socketKeepAlive`: (boolean) Whether to send periodic keep-alive
packets on connected sockets. Defaults to ``False`` (do not send
keep-alive packets).
| **Write Concern options:**
| (Only set if passed. No default values.)
- `w`: (integer or string) If this is a replica set, write operations
will block until they have been replicated to the specified number
or tagged set of servers. `w=<int>` always includes the replica set
primary (e.g. w=3 means write to the primary and wait until
replicated to **two** secondaries). Passing w=0 **disables write
acknowledgement** and all other write concern options.
- `wtimeout`: (integer) Used in conjunction with `w`. Specify a value
in milliseconds to control how long to wait for write propagation
to complete. If replication does not complete in the given
timeframe, a timeout exception is raised.
- `j`: If ``True`` block until write operations have been committed
to the journal. Cannot be used in combination with `fsync`. Prior
to MongoDB 2.6 this option was ignored if the server was running
without journaling. Starting with MongoDB 2.6 write operations will
fail with an exception if this option is used when the server is
running without journaling.
- `fsync`: If ``True`` and the server is running without journaling,
blocks until the server has synced all data files to disk. If the
server is running with journaling, this acts the same as the `j`
option, blocking until write operations have been committed to the
journal. Cannot be used in combination with `j`.
| **Replica set keyword arguments for connecting with a replica set
- either directly or via a mongos:**
- `replicaSet`: (string or None) The name of the replica set to
connect to. The driver will verify that all servers it connects to
match this name. Implies that the hosts specified are a seed list
and the driver should attempt to find all members of the set.
Defaults to ``None``.
- `read_preference`: The read preference for this client. If
connecting directly to a secondary then a read preference mode
*other* than PRIMARY is required - otherwise all queries will throw
:class:`~pymongo.errors.AutoReconnect` "not master".
See :class:`~pymongo.read_preferences.ReadPreference` for all
available read preference options. Defaults to ``PRIMARY``.
| **SSL configuration:**
- `ssl`: If ``True``, create the connection to the server using SSL.
Defaults to ``False``.
- `ssl_keyfile`: The private keyfile used to identify the local
connection against mongod. If included with the ``certfile`` then
only the ``ssl_certfile`` is needed. Implies ``ssl=True``.
Defaults to ``None``.
- `ssl_certfile`: The certificate file used to identify the local
connection against mongod. Implies ``ssl=True``. Defaults to
``None``.
- `ssl_cert_reqs`: Specifies whether a certificate is required from
the other side of the connection, and whether it will be validated
if provided. It must be one of the three values ``ssl.CERT_NONE``
(certificates ignored), ``ssl.CERT_OPTIONAL``
(not required, but validated if provided), or ``ssl.CERT_REQUIRED``
(required and validated). If the value of this parameter is not
``ssl.CERT_NONE`` and a value is not provided for ``ssl_ca_certs``
PyMongo will attempt to load system provided CA certificates.
If the python version in use does not support loading system CA
certificates then the ``ssl_ca_certs`` parameter must point
to a file of CA certificates. Implies ``ssl=True``. Defaults to
``ssl.CERT_REQUIRED`` if not provided and ``ssl=True``.
- `ssl_ca_certs`: The ca_certs file contains a set of concatenated
"certification authority" certificates, which are used to validate
certificates passed from the other end of the connection.
Implies ``ssl=True``. Defaults to ``None``.
- `ssl_match_hostname`: If ``True`` (the default), and
`ssl_cert_reqs` is not ``ssl.CERT_NONE``, enables hostname
verification using the :func:`~ssl.match_hostname` function from
python's :mod:`~ssl` module. Think very carefully before setting
this to ``False`` as that could make your application vulnerable to
man-in-the-middle attacks.
.. mongodoc:: connections
.. versionchanged:: 3.0
:class:`~pymongo.mongo_client.MongoClient` is now the one and only
client class for a standalone server, mongos, or replica set.
It includes the functionality that had been split into
:class:`~pymongo.mongo_client.MongoReplicaSetClient`: it can connect
to a replica set, discover all its members, and monitor the set for
stepdowns, elections, and reconfigs.
The :class:`~pymongo.mongo_client.MongoClient` constructor no
longer blocks while connecting to the server or servers, and it no
longer raises :class:`~pymongo.errors.ConnectionFailure` if they
are unavailable, nor :class:`~pymongo.errors.ConfigurationError`
if the user's credentials are wrong. Instead, the constructor
returns immediately and launches the connection process on
background threads.
Therefore the ``alive`` method is removed since it no longer
provides meaningful information; even if the client is disconnected,
it may discover a server in time to fulfill the next operation.
In PyMongo 2.x, :class:`~pymongo.MongoClient` accepted a list of
standalone MongoDB servers and used the first it could connect to::
MongoClient(['host1.com:27017', 'host2.com:27017'])
A list of multiple standalones is no longer supported; if multiple
servers are listed they must be members of the same replica set, or
mongoses in the same sharded cluster.
The behavior for a list of mongoses is changed from "high
availability" to "load balancing". Before, the client connected to
the lowest-latency mongos in the list, and used it until a network
error prompted it to re-evaluate all mongoses' latencies and
reconnect to one of them. In PyMongo 3, the client monitors its
network latency to all the mongoses continuously, and distributes
operations evenly among those with the lowest latency. See
:ref:`mongos-load-balancing` for more information.
The ``connect`` option is added.
The ``start_request``, ``in_request``, and ``end_request`` methods
are removed, as well as the ``auto_start_request`` option.
The ``copy_database`` method is removed, see the
:doc:`copy_database examples </examples/copydb>` for alternatives.
The :meth:`MongoClient.disconnect` method is removed; it was a
synonym for :meth:`~pymongo.MongoClient.close`.
:class:`~pymongo.mongo_client.MongoClient` no longer returns an
instance of :class:`~pymongo.database.Database` for attribute names
with leading underscores. You must use dict-style lookups instead::
client['__my_database__']
Not::
client.__my_database__
"""
if host is None:
host = self.HOST
if isinstance(host, string_type):
host = [host]
if port is None:
port = self.PORT
if not isinstance(port, int):
raise TypeError("port must be an instance of int")
seeds = set()
username = None
password = None
dbase = None
opts = {}
for entity in host:
if "://" in entity:
if entity.startswith("mongodb://"):
res = uri_parser.parse_uri(entity, port, warn=True)
seeds.update(res["nodelist"])
username = res["username"] or username
password = res["password"] or password
dbase = res["database"] or dbase
opts = res["options"]
else:
idx = entity.find("://")
raise InvalidURI("Invalid URI scheme: "
"%s" % (entity[:idx], ))
else:
seeds.update(uri_parser.split_hosts(entity, port))
if not seeds:
raise ConfigurationError("need to specify at least one host")
# _pool_class, _monitor_class, and _condition_class are for deep
# customization of PyMongo, e.g. Motor.
pool_class = kwargs.pop('_pool_class', None)
monitor_class = kwargs.pop('_monitor_class', None)
condition_class = kwargs.pop('_condition_class', None)
keyword_opts = kwargs
keyword_opts['document_class'] = document_class
keyword_opts['tz_aware'] = tz_aware
keyword_opts['connect'] = connect
# Validate all keyword options.
keyword_opts = dict(
common.validate(k, v) for k, v in keyword_opts.items())
opts.update(keyword_opts)
self.__options = options = ClientOptions(username, password, dbase,
opts)
self.__default_database_name = dbase
self.__lock = threading.Lock()
self.__cursor_manager = CursorManager(self)
self.__kill_cursors_queue = []
# Cache of existing indexes used by ensure_index ops.
self.__index_cache = {}
super(MongoClient,
self).__init__(options.codec_options, options.read_preference,
options.write_concern)
self.__all_credentials = {}
creds = options.credentials
if creds:
self._cache_credentials(creds.source, creds)
self._topology_settings = TopologySettings(
seeds=seeds,
replica_set_name=options.replica_set_name,
pool_class=pool_class,
pool_options=options.pool_options,
monitor_class=monitor_class,
condition_class=condition_class,
local_threshold_ms=options.local_threshold_ms,
server_selection_timeout=options.server_selection_timeout)
self._topology = Topology(self._topology_settings)
if connect:
self._topology.open()
def target():
client = self_ref()
if client is None:
return False # Stop the executor.
MongoClient._process_kill_cursors_queue(client)
return True
executor = periodic_executor.PeriodicExecutor(
condition_class=self._topology_settings.condition_class,
interval=common.KILL_CURSOR_FREQUENCY,
min_interval=0,
target=target,
name="pymongo_kill_cursors_thread")
# We strongly reference the executor and it weakly references us via
# this closure. When the client is freed, stop the executor soon.
self_ref = weakref.ref(self, executor.close)
self._kill_cursors_executor = executor
executor.open()
def _cache_credentials(self, source, credentials, connect=False):
"""Save a set of authentication credentials.
The credentials are used to login a socket whenever one is created.
If `connect` is True, verify the credentials on the server first.
"""
# Don't let other threads affect this call's data.
all_credentials = self.__all_credentials.copy()
if source in all_credentials:
# Nothing to do if we already have these credentials.
if credentials == all_credentials[source]:
return
raise OperationFailure('Another user is already authenticated '
'to this database. You must logout first.')
if connect:
server = self._get_topology().select_server(
writable_preferred_server_selector)
# get_socket() logs out of the database if logged in with old
# credentials, and logs in with new ones.
with server.get_socket(all_credentials) as sock_info:
sock_info.authenticate(credentials)
# If several threads run _cache_credentials at once, last one wins.
self.__all_credentials[source] = credentials
def _purge_credentials(self, source):
"""Purge credentials from the authentication cache."""
self.__all_credentials.pop(source, None)
def _cached(self, dbname, coll, index):
"""Test if `index` is cached."""
cache = self.__index_cache
now = datetime.datetime.utcnow()
return (dbname in cache and coll in cache[dbname]
and index in cache[dbname][coll]
and now < cache[dbname][coll][index])
def _cache_index(self, dbname, collection, index, cache_for):
"""Add an index to the index cache for ensure_index operations."""
now = datetime.datetime.utcnow()
expire = datetime.timedelta(seconds=cache_for) + now
if database not in self.__index_cache:
self.__index_cache[dbname] = {}
self.__index_cache[dbname][collection] = {}
self.__index_cache[dbname][collection][index] = expire
elif collection not in self.__index_cache[dbname]:
self.__index_cache[dbname][collection] = {}
self.__index_cache[dbname][collection][index] = expire
else:
self.__index_cache[dbname][collection][index] = expire
def _purge_index(self,
database_name,
collection_name=None,
index_name=None):
"""Purge an index from the index cache.
If `index_name` is None purge an entire collection.
If `collection_name` is None purge an entire database.
"""
if not database_name in self.__index_cache:
return
if collection_name is None:
del self.__index_cache[database_name]
return
if not collection_name in self.__index_cache[database_name]:
return
if index_name is None:
del self.__index_cache[database_name][collection_name]
return
if index_name in self.__index_cache[database_name][collection_name]:
del self.__index_cache[database_name][collection_name][index_name]
def _server_property(self, attr_name, default=None):
"""An attribute of the current server's description.
Returns "default" while there is no current server, primary, or mongos.
Not threadsafe if used multiple times in a single method, since
the server may change. In such cases, store a local reference to a
ServerDescription first, then use its properties.
"""
try:
server = self._topology.select_server(writable_server_selector,
server_selection_timeout=0)
return getattr(server.description, attr_name)
except ConnectionFailure:
return default
@property
def address(self):
"""(host, port) of the current standalone, primary, or mongos, or None.
Accessing :attr:`address` raises :exc:`~.errors.InvalidOperation` if
the client is load-balancing among mongoses, since there is no single
address. Use :attr:`nodes` instead.
.. versionadded:: 3.0
"""
try:
return self._topology.get_direct_or_primary()
except InvalidOperation:
# Only one case where Topology throws InvalidOperation.
raise InvalidOperation(
'Cannot use "address" property when load balancing among'
' mongoses, use "nodes" instead.')
@property
def primary(self):
"""The (host, port) of the current primary of the replica set.
Returns ``None`` if this client is not connected to a replica set,
there is no primary, or this client was created without the
`replicaSet` option.
.. versionadded:: 3.0
MongoClient gained this property in version 3.0 when
MongoReplicaSetClient's functionality was merged in.
"""
return self._topology.get_primary()
@property
def secondaries(self):
"""The secondary members known to this client.
A sequence of (host, port) pairs. Empty if this client is not
connected to a replica set, there are no visible secondaries, or this
client was created without the `replicaSet` option.
.. versionadded:: 3.0
MongoClient gained this property in version 3.0 when
MongoReplicaSetClient's functionality was merged in.
"""
return self._topology.get_secondaries()
@property
def arbiters(self):
"""Arbiters in the replica set.
A sequence of (host, port) pairs. Empty if this client is not
connected to a replica set, there are no arbiters, or this client was
created without the `replicaSet` option.
"""
return self._topology.get_arbiters()
@property
def is_primary(self):
"""If this client if connected to a server that can accept writes.
True if the current server is a standalone, mongos, or the primary of
a replica set.
"""
return self._server_property('is_writable', False)
@property
def is_mongos(self):
"""If this client is connected to mongos.
"""
return self._server_property('server_type') == SERVER_TYPE.Mongos
@property
def max_pool_size(self):
"""The maximum number of sockets the pool will open concurrently.
When the pool has reached `max_pool_size`, operations block waiting for
a socket to be returned to the pool. If ``waitQueueTimeoutMS`` is set,
a blocked operation will raise :exc:`~pymongo.errors.ConnectionFailure`
after a timeout. By default ``waitQueueTimeoutMS`` is not set.
"""
return self.__options.pool_options.max_pool_size
@property
def nodes(self):
"""Set of all currently connected servers.
.. warning:: When connected to a replica set the value of :attr:`nodes`
can change over time as :class:`MongoClient`'s view of the replica
set changes. :attr:`nodes` can also be an empty set when
:class:`MongoClient` is first instantiated and hasn't yet connected
to any servers, or a network partition causes it to lose connection
to all servers.
"""
description = self._topology.description
return frozenset(s.address for s in description.known_servers)
@property
def max_bson_size(self):
"""The largest BSON object the connected server accepts in bytes.
Defaults to 16MB if not connected to a server.
"""
return self._server_property('max_bson_size', common.MAX_BSON_SIZE)
@property
def max_message_size(self):
"""The largest message the connected server accepts in bytes.
Defaults to 32MB if not connected to a server.
"""
return self._server_property('max_message_size',
common.MAX_MESSAGE_SIZE)
@property
def max_write_batch_size(self):
"""The maxWriteBatchSize reported by the server.
Returns a default value when connected to server versions prior to
MongoDB 2.6.
"""
return self._server_property('max_write_batch_size',
common.MAX_WRITE_BATCH_SIZE)
@property
def local_threshold_ms(self):
"""The local threshold for this instance."""
return self.__options.local_threshold_ms
@property
def server_selection_timeout(self):
"""The server selection timeout for this instance in seconds."""
return self.__options.server_selection_timeout
def _is_writable(self):
"""Attempt to connect to a writable server, or return False.
"""
topology = self._get_topology() # Starts monitors if necessary.
try:
svr = topology.select_server(writable_server_selector)
# When directly connected to a secondary, arbiter, etc.,
# select_server returns it, whatever the selector. Check
# again if the server is writable.
return svr.description.is_writable
except ConnectionFailure:
return False
def close(self):
"""Disconnect from MongoDB.
Close all sockets in the connection pools and stop the monitor threads.
If this instance is used again it will be automatically re-opened and
the threads restarted.
"""
self._topology.close()
def set_cursor_manager(self, manager_class):
"""Set this client's cursor manager.
Raises :class:`TypeError` if `manager_class` is not a subclass of
:class:`~pymongo.cursor_manager.CursorManager`. A cursor manager
handles closing cursors. Different managers can implement different
policies in terms of when to actually kill a cursor that has
been closed.
:Parameters:
- `manager_class`: cursor manager to use
.. versionchanged:: 3.0
Undeprecated.
"""
manager = manager_class(self)
if not isinstance(manager, CursorManager):
raise TypeError("manager_class must be a subclass of "
"CursorManager")
self.__cursor_manager = manager
def _get_topology(self):
"""Get the internal :class:`~pymongo.topology.Topology` object.
If this client was created with "connect=False", calling _get_topology
launches the connection process in the background.
"""
self._topology.open()
return self._topology
@contextlib.contextmanager
def _get_socket(self, selector):
server = self._get_topology().select_server(selector)
try:
with server.get_socket(self.__all_credentials) as sock_info:
yield sock_info
except NetworkTimeout:
# The socket has been closed. Don't reset the server.
# Server Discovery And Monitoring Spec: "When an application
# operation fails because of any network error besides a socket
# timeout...."
raise
except NotMasterError:
# "When the client sees a "not master" error it MUST replace the
# server's description with type Unknown. It MUST request an
# immediate check of the server."
self._reset_server_and_request_check(server.description.address)
raise
except ConnectionFailure:
# "Client MUST replace the server's description with type Unknown
# ... MUST NOT request an immediate check of the server."
self.__reset_server(server.description.address)
raise
def _socket_for_writes(self):
return self._get_socket(writable_server_selector)
@contextlib.contextmanager
def _socket_for_reads(self, read_preference):
preference = read_preference or ReadPreference.PRIMARY
# Get a socket for a server matching the read preference, and yield
# sock_info, slave_ok. Server Selection Spec: "slaveOK must be sent to
# mongods with topology type Single. If the server type is Mongos,
# follow the rules for passing read preference to mongos, even for
# topology type Single."
# Thread safe: if the type is single it cannot change.
topology = self._get_topology()
single = topology.description.topology_type == TOPOLOGY_TYPE.Single
with self._get_socket(read_preference) as sock_info:
slave_ok = (single and not sock_info.is_mongos) or (
preference != ReadPreference.PRIMARY)
yield sock_info, slave_ok
def _send_message_with_response(self,
operation,
read_preference=None,
exhaust=False,
address=None):
"""Send a message to MongoDB and return a Response.
:Parameters:
- `operation`: a _Query or _GetMore object.
- `read_preference` (optional): A ReadPreference.
- `exhaust` (optional): If True, the socket used stays checked out.
It is returned along with its Pool in the Response.
- `address` (optional): Optional address when sending a message
to a specific server, used for getMore.
"""
with self.__lock:
# If needed, restart kill-cursors thread after a fork.
self._kill_cursors_executor.open()
topology = self._get_topology()
if address:
server = topology.select_server_by_address(address)
if not server:
raise AutoReconnect('server %s:%d no longer available' %
address)
else:
selector = read_preference or writable_server_selector
server = topology.select_server(selector)
# A _Query's slaveOk bit is already set for queries with non-primary
# read preference. If this is a direct connection to a mongod, override
# and *always* set the slaveOk bit. See bullet point 2 in
# server-selection.rst#topology-type-single.
set_slave_ok = (
topology.description.topology_type == TOPOLOGY_TYPE.Single
and server.description.server_type != SERVER_TYPE.Mongos)
return self._reset_on_error(server, server.send_message_with_response,
operation, set_slave_ok,
self.__all_credentials, exhaust)
def _reset_on_error(self, server, func, *args, **kwargs):
"""Execute an operation. Reset the server on network error.
Returns fn()'s return value on success. On error, clears the server's
pool and marks the server Unknown.
Re-raises any exception thrown by fn().
"""
try:
return func(*args, **kwargs)
except NetworkTimeout:
# The socket has been closed. Don't reset the server.
raise
except ConnectionFailure:
self.__reset_server(server.description.address)
raise
def __reset_server(self, address):
"""Clear our connection pool for a server and mark it Unknown."""
self._topology.reset_server(address)
def _reset_server_and_request_check(self, address):
"""Clear our pool for a server, mark it Unknown, and check it soon."""
self._topology.reset_server_and_request_check(address)
def __eq__(self, other):
if isinstance(other, self.__class__):
return self.address == other.address
return NotImplemented
def __ne__(self, other):
return not self == other
def _repr_helper(self):
def option_repr(option, value):
"""Fix options whose __repr__ isn't usable in a constructor."""
if option == 'document_class':
if value is dict:
return 'document_class=dict'
else:
return 'document_class=%s.%s' % (value.__module__,
value.__name__)
if "ms" in option:
return "%s='%s'" % (option, int(value * 1000))
return '%s=%r' % (option, value)
# Host first...
options = [
'host=%r' % [
'%s:%d' % (host, port)
for host, port in self._topology_settings.seeds
]
]
# ... then everything in self._constructor_args...
options.extend(
option_repr(key, self.__options._options[key])
for key in self._constructor_args)
# ... then everything else.
options.extend(
option_repr(key, self.__options._options[key])
for key in self.__options._options
if key not in set(self._constructor_args))
return ', '.join(options)
def __repr__(self):
return ("MongoClient(%s)" % (self._repr_helper(), ))
def __getattr__(self, name):
"""Get a database by name.
Raises :class:`~pymongo.errors.InvalidName` if an invalid
database name is used.
:Parameters:
- `name`: the name of the database to get
"""
if name.startswith('_'):
raise AttributeError(
"MongoClient has no attribute %r. To access the %s"
" database, use client[%r]." % (name, name, name))
return self.__getitem__(name)
def __getitem__(self, name):
"""Get a database by name.
Raises :class:`~pymongo.errors.InvalidName` if an invalid
database name is used.
:Parameters:
- `name`: the name of the database to get
"""
return database.Database(self, name)
def close_cursor(self, cursor_id, address=None):
"""Close a single database cursor.
Raises :class:`TypeError` if `cursor_id` is not an instance of
``(int, long)``. What closing the cursor actually means
depends on this client's cursor manager.
:Parameters:
- `cursor_id`: id of cursor to close
- `address` (optional): (host, port) pair of the cursor's server.
If it is not provided, the client attempts to close the cursor on
the primary or standalone, or a mongos server.
.. versionchanged:: 3.0
Added ``address`` parameter.
"""
if not isinstance(cursor_id, integer_types):
raise TypeError("cursor_id must be an instance of (int, long)")
self.__cursor_manager.close(cursor_id, address)
def kill_cursors(self, cursor_ids, address=None):
"""Send a kill cursors message soon with the given ids.
Raises :class:`TypeError` if `cursor_ids` is not an instance of
``list``.
This method may be called from a :class:`~pymongo.cursor.Cursor`
destructor during garbage collection, so it isn't safe to take a
lock or do network I/O. Instead, we schedule the cursor to be closed
soon on a background thread.
:Parameters:
- `cursor_ids`: list of cursor ids to kill
- `address` (optional): (host, port) pair of the cursor's server.
If it is not provided, the client attempts to close the cursor on
the primary or standalone, or a mongos server.
.. versionchanged:: 3.0
Now accepts an `address` argument. Schedules the cursors to be
closed on a background thread instead of sending the message
immediately.
"""
if not isinstance(cursor_ids, list):
raise TypeError("cursor_ids must be a list")
# "Atomic", needs no lock.
self.__kill_cursors_queue.append((address, cursor_ids))
# This method is run periodically by a background thread.
def _process_kill_cursors_queue(self):
"""Process any pending kill cursors requests."""
address_to_cursor_ids = defaultdict(list)
# Other threads or the GC may append to the queue concurrently.
while True:
try:
address, cursor_ids = self.__kill_cursors_queue.pop()
except IndexError:
break
address_to_cursor_ids[address].extend(cursor_ids)
# Don't re-open topology if it's closed and there's no pending cursors.
if address_to_cursor_ids:
publish = monitoring.enabled()
topology = self._get_topology()
for address, cursor_ids in address_to_cursor_ids.items():
try:
if address:
# address could be a tuple or _CursorAddress, but
# select_server_by_address needs (host, port).
server = topology.select_server_by_address(
tuple(address))
else:
# Application called close_cursor() with no address.
server = topology.select_server(
writable_server_selector)
if publish:
start = datetime.datetime.now()
data = message.kill_cursors(cursor_ids)
if publish:
duration = datetime.datetime.now() - start
try:
dbname, collname = address.namespace.split(".", 1)
except AttributeError:
dbname = collname = 'OP_KILL_CURSORS'
command = SON([('killCursors', collname),
('cursors', cursor_ids)])
monitoring.publish_command_start(
command, dbname, data[0], address)
start = datetime.datetime.now()
server.send_message(data, self.__all_credentials)
if publish:
duration = (datetime.datetime.now() - start) + duration
# OP_KILL_CURSORS returns no reply, fake one.
reply = {'cursorsUnknown': cursor_ids, 'ok': 1}
monitoring.publish_command_success(
duration, reply, 'killCursors', data[0], address)
except ConnectionFailure as exc:
warnings.warn("couldn't close cursor on %s: %s" %
(address, exc))
def server_info(self):
"""Get information about the MongoDB server we're connected to."""
return self.admin.command("buildinfo",
read_preference=ReadPreference.PRIMARY)
def database_names(self):
"""Get a list of the names of all databases on the connected server."""
return [
db["name"] for db in self.admin.command(
"listDatabases", read_preference=ReadPreference.PRIMARY)
["databases"]
]
def drop_database(self, name_or_database):
"""Drop a database.
Raises :class:`TypeError` if `name_or_database` is not an instance of
:class:`basestring` (:class:`str` in python 3) or
:class:`~pymongo.database.Database`.
:Parameters:
- `name_or_database`: the name of a database to drop, or a
:class:`~pymongo.database.Database` instance representing the
database to drop
"""
name = name_or_database
if isinstance(name, database.Database):
name = name.name
if not isinstance(name, string_type):
raise TypeError("name_or_database must be an instance "
"of %s or a Database" % (string_type.__name__, ))
self._purge_index(name)
self[name].command("dropDatabase",
read_preference=ReadPreference.PRIMARY)
def get_default_database(self):
"""Get the database named in the MongoDB connection URI.
>>> uri = 'mongodb://host/my_database'
>>> client = MongoClient(uri)
>>> db = client.get_default_database()
>>> assert db.name == 'my_database'
Useful in scripts where you want to choose which database to use
based only on the URI in a configuration file.
"""
if self.__default_database_name is None:
raise ConfigurationError('No default database defined')
return self[self.__default_database_name]
def get_database(self,
name,
codec_options=None,
read_preference=None,
write_concern=None):
"""Get a :class:`~pymongo.database.Database` with the given name and
options.
Useful for creating a :class:`~pymongo.database.Database` with
different codec options, read preference, and/or write concern from
this :class:`MongoClient`.
>>> client.read_preference
Primary()
>>> db1 = client.test
>>> db1.read_preference
Primary()
>>> from pymongo import ReadPreference
>>> db2 = client.get_database(
... 'test', read_preference=ReadPreference.SECONDARY)
>>> db2.read_preference
Secondary(tag_sets=None)
:Parameters:
- `name`: The name of the database - a string.
- `codec_options` (optional): An instance of
:class:`~bson.codec_options.CodecOptions`. If ``None`` (the
default) the :attr:`codec_options` of this :class:`MongoClient` is
used.
- `read_preference` (optional): The read preference to use. If
``None`` (the default) the :attr:`read_preference` of this
:class:`MongoClient` is used. See :mod:`~pymongo.read_preferences`
for options.
- `write_concern` (optional): An instance of
:class:`~pymongo.write_concern.WriteConcern`. If ``None`` (the
default) the :attr:`write_concern` of this :class:`MongoClient` is
used.
"""
return database.Database(self, name, codec_options, read_preference,
write_concern)
@property
def is_locked(self):
"""Is this server locked? While locked, all write operations
are blocked, although read operations may still be allowed.
Use :meth:`unlock` to unlock.
"""
ops = self.admin.current_op()
return bool(ops.get('fsyncLock', 0))
def fsync(self, **kwargs):
"""Flush all pending writes to datafiles.
:Parameters:
Optional parameters can be passed as keyword arguments:
- `lock`: If True lock the server to disallow writes.
- `async`: If True don't block while synchronizing.
.. warning:: `async` and `lock` can not be used together.
.. warning:: MongoDB does not support the `async` option
on Windows and will raise an exception on that
platform.
"""
self.admin.command("fsync",
read_preference=ReadPreference.PRIMARY,
**kwargs)
def unlock(self):
"""Unlock a previously locked server.
"""
coll = self.admin.get_collection(
"$cmd.sys.unlock", read_preference=ReadPreference.PRIMARY)
coll.find_one()
def __enter__(self):
return self
def __exit__(self, exc_type, exc_val, exc_tb):
self.close()
def __iter__(self):
return self
def __next__(self):
raise TypeError("'MongoClient' object is not iterable")
next = __next__
class MongoClient(common.BaseObject):
HOST = "localhost"
PORT = 27017
def __init__(
self,
host=None,
port=None,
document_class=dict,
tz_aware=False,
connect=True,
**kwargs):
"""Client for a MongoDB instance, a replica set, or a set of mongoses.
The client object is thread-safe and has connection-pooling built in.
If an operation fails because of a network error,
:class:`~pymongo.errors.ConnectionFailure` is raised and the client
reconnects in the background. Application code should handle this
exception (recognizing that the operation failed) and then continue to
execute.
The `host` parameter can be a full `mongodb URI
<http://dochub.mongodb.org/core/connections>`_, in addition to
a simple hostname. It can also be a list of hostnames or
URIs. Any port specified in the host string(s) will override
the `port` parameter. If multiple mongodb URIs containing
database or auth information are passed, the last database,
username, and password present will be used. For username and
passwords reserved characters like ':', '/', '+' and '@' must be
escaped following RFC 2396.
:Parameters:
- `host` (optional): hostname or IP address of the
instance to connect to, or a mongodb URI, or a list of
hostnames / mongodb URIs. If `host` is an IPv6 literal
it must be enclosed in '[' and ']' characters following
the RFC2732 URL syntax (e.g. '[::1]' for localhost)
- `port` (optional): port number on which to connect
- `document_class` (optional): default class to use for
documents returned from queries on this client
- `tz_aware` (optional): if ``True``,
:class:`~datetime.datetime` instances returned as values
in a document by this :class:`MongoClient` will be timezone
aware (otherwise they will be naive)
- `connect` (optional): if ``True`` (the default), immediately
begin connecting to MongoDB in the background. Otherwise connect
on the first operation.
| **Other optional parameters can be passed as keyword arguments:**
- `maxPoolSize` (optional): The maximum number of connections
that the pool will open simultaneously. If this is set, operations
will block if there are `maxPoolSize` outstanding connections
from the pool. Defaults to 100. Cannot be 0.
- `socketTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait for a response after sending an
ordinary (non-monitoring) database operation before concluding that
a network error has occurred. Defaults to ``None`` (no timeout).
- `connectTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait during server monitoring when
connecting a new socket to a server before concluding the server
is unavailable. Defaults to ``20000`` (20 seconds).
- `serverSelectionTimeoutMS`: (integer) Controls how long (in
milliseconds) the driver will wait to find an available,
appropriate server to carry out a database operation; while it is
waiting, multiple server monitoring operations may be carried out,
each controlled by `connectTimeoutMS`. Defaults to ``30000`` (30
seconds).
- `waitQueueTimeoutMS`: (integer or None) How long (in milliseconds)
a thread will wait for a socket from the pool if the pool has no
free sockets. Defaults to ``None`` (no timeout).
- `waitQueueMultiple`: (integer or None) Multiplied by maxPoolSize
to give the number of threads allowed to wait for a socket at one
time. Defaults to ``None`` (no limit).
- `socketKeepAlive`: (boolean) Whether to send periodic keep-alive
packets on connected sockets. Defaults to ``False`` (do not send
keep-alive packets).
| **Write Concern options:**
| (Only set if passed. No default values.)
- `w`: (integer or string) If this is a replica set, write operations
will block until they have been replicated to the specified number
or tagged set of servers. `w=<int>` always includes the replica set
primary (e.g. w=3 means write to the primary and wait until
replicated to **two** secondaries). Passing w=0 **disables write
acknowledgement** and all other write concern options.
- `wtimeout`: (integer) Used in conjunction with `w`. Specify a value
in milliseconds to control how long to wait for write propagation
to complete. If replication does not complete in the given
timeframe, a timeout exception is raised.
- `j`: If ``True`` block until write operations have been committed
to the journal. Cannot be used in combination with `fsync`. Prior
to MongoDB 2.6 this option was ignored if the server was running
without journaling. Starting with MongoDB 2.6 write operations will
fail with an exception if this option is used when the server is
running without journaling.
- `fsync`: If ``True`` and the server is running without journaling,
blocks until the server has synced all data files to disk. If the
server is running with journaling, this acts the same as the `j`
option, blocking until write operations have been committed to the
journal. Cannot be used in combination with `j`.
| **Replica set keyword arguments for connecting with a replica set
- either directly or via a mongos:**
- `replicaSet`: (string or None) The name of the replica set to
connect to. The driver will verify that all servers it connects to
match this name. Implies that the hosts specified are a seed list
and the driver should attempt to find all members of the set.
Defaults to ``None``.
- `read_preference`: The read preference for this client. If
connecting directly to a secondary then a read preference mode
*other* than PRIMARY is required - otherwise all queries will throw
:class:`~pymongo.errors.AutoReconnect` "not master".
See :class:`~pymongo.read_preferences.ReadPreference` for all
available read preference options. Defaults to ``PRIMARY``.
| **SSL configuration:**
- `ssl`: If ``True``, create the connection to the server using SSL.
Defaults to ``False``.
- `ssl_keyfile`: The private keyfile used to identify the local
connection against mongod. If included with the ``certfile`` then
only the ``ssl_certfile`` is needed. Implies ``ssl=True``.
Defaults to ``None``.
- `ssl_certfile`: The certificate file used to identify the local
connection against mongod. Implies ``ssl=True``. Defaults to
``None``.
- `ssl_cert_reqs`: Specifies whether a certificate is required from
the other side of the connection, and whether it will be validated
if provided. It must be one of the three values ``ssl.CERT_NONE``
(certificates ignored), ``ssl.CERT_OPTIONAL``
(not required, but validated if provided), or ``ssl.CERT_REQUIRED``
(required and validated). If the value of this parameter is not
``ssl.CERT_NONE`` and a value is not provided for ``ssl_ca_certs``
PyMongo will attempt to load system provided CA certificates.
If the python version in use does not support loading system CA
certificates then the ``ssl_ca_certs`` parameter must point
to a file of CA certificates. Implies ``ssl=True``. Defaults to
``ssl.CERT_REQUIRED`` if not provided and ``ssl=True``.
- `ssl_ca_certs`: The ca_certs file contains a set of concatenated
"certification authority" certificates, which are used to validate
certificates passed from the other end of the connection.
Implies ``ssl=True``. Defaults to ``None``.
- `ssl_match_hostname`: If ``True`` (the default), and
`ssl_cert_reqs` is not ``ssl.CERT_NONE``, enables hostname
verification using the :func:`~ssl.match_hostname` function from
python's :mod:`~ssl` module. Think very carefully before setting
this to ``False`` as that could make your application vulnerable to
man-in-the-middle attacks.
.. mongodoc:: connections
.. versionchanged:: 3.0
:class:`~pymongo.mongo_client.MongoClient` is now the one and only
client class for a standalone server, mongos, or replica set.
It includes the functionality that had been split into
:class:`~pymongo.mongo_client.MongoReplicaSetClient`: it can connect
to a replica set, discover all its members, and monitor the set for
stepdowns, elections, and reconfigs.
The :class:`~pymongo.mongo_client.MongoClient` constructor no
longer blocks while connecting to the server or servers, and it no
longer raises :class:`~pymongo.errors.ConnectionFailure` if they
are unavailable, nor :class:`~pymongo.errors.ConfigurationError`
if the user's credentials are wrong. Instead, the constructor
returns immediately and launches the connection process on
background threads.
Therefore the ``alive`` method is removed since it no longer
provides meaningful information; even if the client is disconnected,
it may discover a server in time to fulfill the next operation.
In PyMongo 2.x, :class:`~pymongo.MongoClient` accepted a list of
standalone MongoDB servers and used the first it could connect to::
MongoClient(['host1.com:27017', 'host2.com:27017'])
A list of multiple standalones is no longer supported; if multiple
servers are listed they must be members of the same replica set, or
mongoses in the same sharded cluster.
The behavior for a list of mongoses is changed from "high
availability" to "load balancing". Before, the client connected to
the lowest-latency mongos in the list, and used it until a network
error prompted it to re-evaluate all mongoses' latencies and
reconnect to one of them. In PyMongo 3, the client monitors its
network latency to all the mongoses continuously, and distributes
operations evenly among those with the lowest latency. See
:ref:`mongos-load-balancing` for more information.
The ``connect`` option is added.
The ``start_request``, ``in_request``, and ``end_request`` methods
are removed, as well as the ``auto_start_request`` option.
The ``copy_database`` method is removed, see the
:doc:`copy_database examples </examples/copydb>` for alternatives.
The :meth:`MongoClient.disconnect` method is removed; it was a
synonym for :meth:`~pymongo.MongoClient.close`.
:class:`~pymongo.mongo_client.MongoClient` no longer returns an
instance of :class:`~pymongo.database.Database` for attribute names
with leading underscores. You must use dict-style lookups instead::
client['__my_database__']
Not::
client.__my_database__
"""
if host is None:
host = self.HOST
if isinstance(host, string_type):
host = [host]
if port is None:
port = self.PORT
if not isinstance(port, int):
raise TypeError("port must be an instance of int")
seeds = set()
username = None
password = None
dbase = None
opts = {}
for entity in host:
if "://" in entity:
if entity.startswith("mongodb://"):
res = uri_parser.parse_uri(entity, port, False)
seeds.update(res["nodelist"])
username = res["username"] or username
password = res["password"] or password
dbase = res["database"] or dbase
opts = res["options"]
else:
idx = entity.find("://")
raise InvalidURI("Invalid URI scheme: "
"%s" % (entity[:idx],))
else:
seeds.update(uri_parser.split_hosts(entity, port))
if not seeds:
raise ConfigurationError("need to specify at least one host")
# _pool_class, _monitor_class, and _condition_class are for deep
# customization of PyMongo, e.g. Motor.
pool_class = kwargs.pop('_pool_class', None)
monitor_class = kwargs.pop('_monitor_class', None)
condition_class = kwargs.pop('_condition_class', None)
kwargs['document_class'] = document_class
kwargs['tz_aware'] = tz_aware
opts.update(kwargs)
self.__options = options = ClientOptions(
username, password, dbase, opts)
self.__default_database_name = dbase
self.__lock = threading.Lock()
self.__cursor_manager = CursorManager(self)
self.__kill_cursors_queue = []
# Cache of existing indexes used by ensure_index ops.
self.__index_cache = {}
super(MongoClient, self).__init__(options.codec_options,
options.read_preference,
options.write_concern)
self.__all_credentials = {}
creds = options.credentials
if creds:
self._cache_credentials(creds.source, creds)
self._topology_settings = TopologySettings(
seeds=seeds,
replica_set_name=options.replica_set_name,
pool_class=pool_class,
pool_options=options.pool_options,
monitor_class=monitor_class,
condition_class=condition_class,
local_threshold_ms=options.local_threshold_ms,
server_selection_timeout=options.server_selection_timeout)
self._topology = Topology(self._topology_settings)
if connect:
self._topology.open()
def target():
client = self_ref()
if client is None:
return False # Stop the executor.
MongoClient._process_kill_cursors_queue(client)
return True
executor = periodic_executor.PeriodicExecutor(
condition_class=self._topology_settings.condition_class,
interval=common.KILL_CURSOR_FREQUENCY,
min_interval=0,
target=target)
# We strongly reference the executor and it weakly references us via
# this closure. When the client is freed, stop the executor soon.
self_ref = weakref.ref(self, executor.close)
self._kill_cursors_executor = executor
executor.open()
def _cache_credentials(self, source, credentials, connect=False):
"""Save a set of authentication credentials.
The credentials are used to login a socket whenever one is created.
If `connect` is True, verify the credentials on the server first.
"""
# Don't let other threads affect this call's data.
all_credentials = self.__all_credentials.copy()
if source in all_credentials:
# Nothing to do if we already have these credentials.
if credentials == all_credentials[source]:
return
raise OperationFailure('Another user is already authenticated '
'to this database. You must logout first.')
if connect:
server = self._get_topology().select_server(
writable_preferred_server_selector)
# get_socket() logs out of the database if logged in with old
# credentials, and logs in with new ones.
with server.get_socket(all_credentials) as sock_info:
sock_info.authenticate(credentials)
# If several threads run _cache_credentials at once, last one wins.
self.__all_credentials[source] = credentials
def _purge_credentials(self, source):
"""Purge credentials from the authentication cache."""
self.__all_credentials.pop(source, None)
def _cached(self, dbname, coll, index):
"""Test if `index` is cached."""
cache = self.__index_cache
now = datetime.datetime.utcnow()
return (dbname in cache and
coll in cache[dbname] and
index in cache[dbname][coll] and
now < cache[dbname][coll][index])
def _cache_index(self, dbname, collection, index, cache_for):
"""Add an index to the index cache for ensure_index operations."""
now = datetime.datetime.utcnow()
expire = datetime.timedelta(seconds=cache_for) + now
if database not in self.__index_cache:
self.__index_cache[dbname] = {}
self.__index_cache[dbname][collection] = {}
self.__index_cache[dbname][collection][index] = expire
elif collection not in self.__index_cache[dbname]:
self.__index_cache[dbname][collection] = {}
self.__index_cache[dbname][collection][index] = expire
else:
self.__index_cache[dbname][collection][index] = expire
def _purge_index(self, database_name,
collection_name=None, index_name=None):
"""Purge an index from the index cache.
If `index_name` is None purge an entire collection.
If `collection_name` is None purge an entire database.
"""
if not database_name in self.__index_cache:
return
if collection_name is None:
del self.__index_cache[database_name]
return
if not collection_name in self.__index_cache[database_name]:
return
if index_name is None:
del self.__index_cache[database_name][collection_name]
return
if index_name in self.__index_cache[database_name][collection_name]:
del self.__index_cache[database_name][collection_name][index_name]
def _server_property(self, attr_name, default=None):
"""An attribute of the current server's description.
Returns "default" while there is no current server, primary, or mongos.
Not threadsafe if used multiple times in a single method, since
the server may change. In such cases, store a local reference to a
ServerDescription first, then use its properties.
"""
try:
server = self._topology.select_server(
writable_server_selector, server_selection_timeout=0)
return getattr(server.description, attr_name)
except ConnectionFailure:
return default
@property
def address(self):
"""(host, port) of the current standalone, primary, or mongos, or None.
Accessing :attr:`address` raises :exc:`~.errors.InvalidOperation` if
the client is load-balancing among mongoses, since there is no single
address. Use :attr:`nodes` instead.
.. versionadded:: 3.0
"""
try:
return self._topology.get_direct_or_primary()
except InvalidOperation:
# Only one case where Topology throws InvalidOperation.
raise InvalidOperation(
'Cannot use "address" property when load balancing among'
' mongoses, use "nodes" instead.')
@property
def primary(self):
"""The (host, port) of the current primary of the replica set.
Returns ``None`` if this client is not connected to a replica set,
there is no primary, or this client was created without the
`replicaSet` option.
.. versionadded:: 3.0
MongoClient gained this property in version 3.0 when
MongoReplicaSetClient's functionality was merged in.
"""
return self._topology.get_primary()
@property
def secondaries(self):
"""The secondary members known to this client.
A sequence of (host, port) pairs. Empty if this client is not
connected to a replica set, there are no visible secondaries, or this
client was created without the `replicaSet` option.
.. versionadded:: 3.0
MongoClient gained this property in version 3.0 when
MongoReplicaSetClient's functionality was merged in.
"""
return self._topology.get_secondaries()
@property
def arbiters(self):
"""Arbiters in the replica set.
A sequence of (host, port) pairs. Empty if this client is not
connected to a replica set, there are no arbiters, or this client was
created without the `replicaSet` option.
"""
return self._topology.get_arbiters()
@property
def is_primary(self):
"""If this client if connected to a server that can accept writes.
True if the current server is a standalone, mongos, or the primary of
a replica set.
"""
return self._server_property('is_writable', False)
@property
def is_mongos(self):
"""If this client is connected to mongos.
"""
return self._server_property('server_type') == SERVER_TYPE.Mongos
@property
def max_pool_size(self):
"""The maximum number of sockets the pool will open concurrently.
When the pool has reached `max_pool_size`, operations block waiting for
a socket to be returned to the pool. If ``waitQueueTimeoutMS`` is set,
a blocked operation will raise :exc:`~pymongo.errors.ConnectionFailure`
after a timeout. By default ``waitQueueTimeoutMS`` is not set.
"""
return self.__options.pool_options.max_pool_size
@property
def nodes(self):
"""Set of all currently connected servers.
.. warning:: When connected to a replica set the value of :attr:`nodes`
can change over time as :class:`MongoClient`'s view of the replica
set changes. :attr:`nodes` can also be an empty set when
:class:`MongoClient` is first instantiated and hasn't yet connected
to any servers, or a network partition causes it to lose connection
to all servers.
"""
description = self._topology.description
return frozenset(s.address for s in description.known_servers)
@property
def max_bson_size(self):
"""The largest BSON object the connected server accepts in bytes.
Defaults to 16MB if not connected to a server.
"""
return self._server_property('max_bson_size', common.MAX_BSON_SIZE)
@property
def max_message_size(self):
"""The largest message the connected server accepts in bytes.
Defaults to 32MB if not connected to a server.
"""
return self._server_property(
'max_message_size', common.MAX_MESSAGE_SIZE)
@property
def max_write_batch_size(self):
"""The maxWriteBatchSize reported by the server.
Returns a default value when connected to server versions prior to
MongoDB 2.6.
"""
return self._server_property(
'max_write_batch_size', common.MAX_WRITE_BATCH_SIZE)
@property
def local_threshold_ms(self):
"""The local threshold for this instance."""
return self.__options.local_threshold_ms
@property
def server_selection_timeout(self):
"""The server selection timeout for this instance in seconds."""
return self.__options.server_selection_timeout
def _is_writable(self):
"""Attempt to connect to a writable server, or return False.
"""
topology = self._get_topology() # Starts monitors if necessary.
try:
svr = topology.select_server(writable_server_selector)
# When directly connected to a secondary, arbiter, etc.,
# select_server returns it, whatever the selector. Check
# again if the server is writable.
return svr.description.is_writable
except ConnectionFailure:
return False
def close(self):
"""Disconnect from MongoDB.
Close all sockets in the connection pools and stop the monitor threads.
If this instance is used again it will be automatically re-opened and
the threads restarted.
"""
self._topology.close()
def set_cursor_manager(self, manager_class):
"""Set this client's cursor manager.
Raises :class:`TypeError` if `manager_class` is not a subclass of
:class:`~pymongo.cursor_manager.CursorManager`. A cursor manager
handles closing cursors. Different managers can implement different
policies in terms of when to actually kill a cursor that has
been closed.
:Parameters:
- `manager_class`: cursor manager to use
.. versionchanged:: 3.0
Undeprecated.
"""
manager = manager_class(self)
if not isinstance(manager, CursorManager):
raise TypeError("manager_class must be a subclass of "
"CursorManager")
self.__cursor_manager = manager
def _get_topology(self):
"""Get the internal :class:`~pymongo.topology.Topology` object.
If this client was created with "connect=False", calling _get_topology
launches the connection process in the background.
"""
self._topology.open()
return self._topology
@contextlib.contextmanager
def _get_socket(self, selector):
server = self._get_topology().select_server(selector)
try:
with server.get_socket(self.__all_credentials) as sock_info:
yield sock_info
except NetworkTimeout:
# The socket has been closed. Don't reset the server.
# Server Discovery And Monitoring Spec: "When an application
# operation fails because of any network error besides a socket
# timeout...."
raise
except NotMasterError:
# "When the client sees a "not master" error it MUST replace the
# server's description with type Unknown. It MUST request an
# immediate check of the server."
self._reset_server_and_request_check(server.description.address)
raise
except ConnectionFailure:
# "Client MUST replace the server's description with type Unknown
# ... MUST NOT request an immediate check of the server."
self.__reset_server(server.description.address)
raise
def _socket_for_writes(self):
return self._get_socket(writable_server_selector)
@contextlib.contextmanager
def _socket_for_reads(self, read_preference):
preference = read_preference or ReadPreference.PRIMARY
# Get a socket for a server matching the read preference, and yield
# sock_info, slave_ok. Server Selection Spec: "slaveOK must be sent to
# mongods with topology type Single. If the server type is Mongos,
# follow the rules for passing read preference to mongos, even for
# topology type Single."
# Thread safe: if the type is single it cannot change.
topology = self._get_topology()
single = topology.description.topology_type == TOPOLOGY_TYPE.Single
with self._get_socket(read_preference) as sock_info:
slave_ok = (single and not sock_info.is_mongos) or (
preference != ReadPreference.PRIMARY)
yield sock_info, slave_ok
def _send_message_with_response(self, operation, read_preference=None,
exhaust=False, address=None):
"""Send a message to MongoDB and return a Response.
:Parameters:
- `operation`: a _Query or _GetMore object.
- `read_preference` (optional): A ReadPreference.
- `exhaust` (optional): If True, the socket used stays checked out.
It is returned along with its Pool in the Response.
- `address` (optional): Optional address when sending a message
to a specific server, used for getMore.
"""
with self.__lock:
# If needed, restart kill-cursors thread after a fork.
self._kill_cursors_executor.open()
topology = self._get_topology()
if address:
server = topology.select_server_by_address(address)
if not server:
raise AutoReconnect('server %s:%d no longer available'
% address)
else:
selector = read_preference or writable_server_selector
server = topology.select_server(selector)
# A _Query's slaveOk bit is already set for queries with non-primary
# read preference. If this is a direct connection to a mongod, override
# and *always* set the slaveOk bit. See bullet point 2 in
# server-selection.rst#topology-type-single.
set_slave_ok = (
topology.description.topology_type == TOPOLOGY_TYPE.Single
and server.description.server_type != SERVER_TYPE.Mongos)
return self._reset_on_error(
server,
server.send_message_with_response,
operation,
set_slave_ok,
self.__all_credentials,
exhaust)
def _reset_on_error(self, server, func, *args, **kwargs):
"""Execute an operation. Reset the server on network error.
Returns fn()'s return value on success. On error, clears the server's
pool and marks the server Unknown.
Re-raises any exception thrown by fn().
"""
try:
return func(*args, **kwargs)
except NetworkTimeout:
# The socket has been closed. Don't reset the server.
raise
except ConnectionFailure:
self.__reset_server(server.description.address)
raise
def __reset_server(self, address):
"""Clear our connection pool for a server and mark it Unknown."""
self._topology.reset_server(address)
def _reset_server_and_request_check(self, address):
"""Clear our pool for a server, mark it Unknown, and check it soon."""
self._topology.reset_server_and_request_check(address)
def __eq__(self, other):
if isinstance(other, self.__class__):
return self.address == other.address
return NotImplemented
def __ne__(self, other):
return not self == other
def __repr__(self):
server_descriptions = self._topology.description.server_descriptions()
if len(server_descriptions) == 1:
description, = server_descriptions.values()
return "MongoClient(%r, %r)" % description.address
else:
return "MongoClient(%r)" % [
"%s:%d" % address for address in server_descriptions]
def __getattr__(self, name):
"""Get a database by name.
Raises :class:`~pymongo.errors.InvalidName` if an invalid
database name is used.
:Parameters:
- `name`: the name of the database to get
"""
if name.startswith('_'):
raise AttributeError(
"MongoClient has no attribute %r. To access the %s"
" database, use client[%r]." % (name, name, name))
return self.__getitem__(name)
def __getitem__(self, name):
"""Get a database by name.
Raises :class:`~pymongo.errors.InvalidName` if an invalid
database name is used.
:Parameters:
- `name`: the name of the database to get
"""
return database.Database(self, name)
def close_cursor(self, cursor_id, address=None):
"""Close a single database cursor.
Raises :class:`TypeError` if `cursor_id` is not an instance of
``(int, long)``. What closing the cursor actually means
depends on this client's cursor manager.
:Parameters:
- `cursor_id`: id of cursor to close
- `address` (optional): (host, port) pair of the cursor's server.
If it is not provided, the client attempts to close the cursor on
the primary or standalone, or a mongos server.
.. versionchanged:: 3.0
Added ``address`` parameter.
"""
if not isinstance(cursor_id, integer_types):
raise TypeError("cursor_id must be an instance of (int, long)")
self.__cursor_manager.close(cursor_id, address)
def kill_cursors(self, cursor_ids, address=None):
"""Send a kill cursors message soon with the given ids.
Raises :class:`TypeError` if `cursor_ids` is not an instance of
``list``.
This method may be called from a :class:`~pymongo.cursor.Cursor`
destructor during garbage collection, so it isn't safe to take a
lock or do network I/O. Instead, we schedule the cursor to be closed
soon on a background thread.
:Parameters:
- `cursor_ids`: list of cursor ids to kill
- `address` (optional): (host, port) pair of the cursor's server.
If it is not provided, the client attempts to close the cursor on
the primary or standalone, or a mongos server.
.. versionchanged:: 3.0
Now accepts an `address` argument. Schedules the cursors to be
closed on a background thread instead of sending the message
immediately.
"""
if not isinstance(cursor_ids, list):
raise TypeError("cursor_ids must be a list")
# "Atomic", needs no lock.
self.__kill_cursors_queue.append((address, cursor_ids))
# This method is run periodically by a background thread.
def _process_kill_cursors_queue(self):
"""Process any pending kill cursors requests."""
address_to_cursor_ids = defaultdict(list)
# Other threads or the GC may append to the queue concurrently.
while True:
try:
address, cursor_ids = self.__kill_cursors_queue.pop()
except IndexError:
break
address_to_cursor_ids[address].extend(cursor_ids)
# Don't re-open topology if it's closed and there's no pending cursors.
if address_to_cursor_ids:
topology = self._get_topology()
for address, cursor_ids in address_to_cursor_ids.items():
try:
if address:
server = topology.select_server_by_address(address)
else:
# Application called close_cursor() with no address.
server = topology.select_server(
writable_server_selector)
server.send_message(message.kill_cursors(cursor_ids),
self.__all_credentials)
except ConnectionFailure as exc:
warnings.warn("couldn't close cursor on %s: %s"
% (address, exc))
def server_info(self):
"""Get information about the MongoDB server we're connected to."""
return self.admin.command("buildinfo",
read_preference=ReadPreference.PRIMARY)
def database_names(self):
"""Get a list of the names of all databases on the connected server."""
return [db["name"] for db in
self.admin.command(
"listDatabases",
read_preference=ReadPreference.PRIMARY)["databases"]]
def drop_database(self, name_or_database):
"""Drop a database.
Raises :class:`TypeError` if `name_or_database` is not an instance of
:class:`basestring` (:class:`str` in python 3) or
:class:`~pymongo.database.Database`.
:Parameters:
- `name_or_database`: the name of a database to drop, or a
:class:`~pymongo.database.Database` instance representing the
database to drop
"""
name = name_or_database
if isinstance(name, database.Database):
name = name.name
if not isinstance(name, string_type):
raise TypeError("name_or_database must be an instance "
"of %s or a Database" % (string_type.__name__,))
self._purge_index(name)
self[name].command("dropDatabase",
read_preference=ReadPreference.PRIMARY)
def get_default_database(self):
"""Get the database named in the MongoDB connection URI.
>>> uri = 'mongodb://host/my_database'
>>> client = MongoClient(uri)
>>> db = client.get_default_database()
>>> assert db.name == 'my_database'
Useful in scripts where you want to choose which database to use
based only on the URI in a configuration file.
"""
if self.__default_database_name is None:
raise ConfigurationError('No default database defined')
return self[self.__default_database_name]
def get_database(self, name, codec_options=None,
read_preference=None, write_concern=None):
"""Get a :class:`~pymongo.database.Database` with the given name and
options.
Useful for creating a :class:`~pymongo.database.Database` with
different codec options, read preference, and/or write concern from
this :class:`MongoClient`.
>>> client.read_preference
Primary()
>>> db1 = client.test
>>> db1.read_preference
Primary()
>>> from pymongo import ReadPreference
>>> db2 = client.get_database(
... 'test', read_preference=ReadPreference.SECONDARY)
>>> db2.read_preference
Secondary(tag_sets=None)
:Parameters:
- `name`: The name of the database - a string.
- `codec_options` (optional): An instance of
:class:`~bson.codec_options.CodecOptions`. If ``None`` (the
default) the :attr:`codec_options` of this :class:`MongoClient` is
used.
- `read_preference` (optional): The read preference to use. If
``None`` (the default) the :attr:`read_preference` of this
:class:`MongoClient` is used. See :mod:`~pymongo.read_preferences`
for options.
- `write_concern` (optional): An instance of
:class:`~pymongo.write_concern.WriteConcern`. If ``None`` (the
default) the :attr:`write_concern` of this :class:`MongoClient` is
used.
"""
return database.Database(
self, name, codec_options, read_preference, write_concern)
@property
def is_locked(self):
"""Is this server locked? While locked, all write operations
are blocked, although read operations may still be allowed.
Use :meth:`unlock` to unlock.
"""
ops = self.admin.current_op()
return bool(ops.get('fsyncLock', 0))
def fsync(self, **kwargs):
"""Flush all pending writes to datafiles.
:Parameters:
Optional parameters can be passed as keyword arguments:
- `lock`: If True lock the server to disallow writes.
- `async`: If True don't block while synchronizing.
.. warning:: `async` and `lock` can not be used together.
.. warning:: MongoDB does not support the `async` option
on Windows and will raise an exception on that
platform.
"""
self.admin.command("fsync",
read_preference=ReadPreference.PRIMARY, **kwargs)
def unlock(self):
"""Unlock a previously locked server.
"""
coll = self.admin.get_collection(
"$cmd.sys.unlock", read_preference=ReadPreference.PRIMARY)
coll.find_one()
def __enter__(self):
return self
def __exit__(self, exc_type, exc_val, exc_tb):
self.close()
def __iter__(self):
return self
def __next__(self):
raise TypeError("'MongoClient' object is not iterable")
next = __next__
def run_scenario(self):
if 'heartbeatFrequencyMS' in scenario_def:
frequency = int(scenario_def['heartbeatFrequencyMS']) / 1000.0
else:
frequency = HEARTBEAT_FREQUENCY
# Initialize topologies.
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
topology = Topology(
TopologySettings(seeds=seeds,
monitor_class=MockMonitor,
pool_class=MockPool,
heartbeat_frequency=frequency))
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
topology.on_change(server_description)
# Create server selector.
# Make first letter lowercase to match read_pref's modes.
pref_def = scenario_def['read_preference']
mode_string = pref_def.get('mode', 'primary')
mode_string = mode_string[:1].lower() + mode_string[1:]
mode = read_preferences.read_pref_mode_from_name(mode_string)
max_staleness = pref_def.get('maxStalenessMS', 0) / 1000.0
if not max_staleness:
max_staleness = None
tag_sets = pref_def.get('tag_sets')
if scenario_def.get('error'):
with self.assertRaises(ConfigurationError):
# Error can be raised when making Read Pref or selecting.
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
topology.select_server(pref)
return
expected_addrs = set([
server['address'] for server in scenario_def['in_latency_window']])
# Select servers.
pref = read_preferences.make_read_preference(
mode, tag_sets=tag_sets, max_staleness=max_staleness)
if not expected_addrs:
with self.assertRaises(ConnectionFailure):
topology.select_servers(pref, server_selection_timeout=0)
return
servers = topology.select_servers(pref, server_selection_timeout=0)
actual_addrs = set(['%s:%d' % s.description.address for s in servers])
for unexpected in actual_addrs - expected_addrs:
self.fail("'%s' shouldn't have been selected, but was" % unexpected)
for unselected in expected_addrs - actual_addrs:
self.fail("'%s' should have been selected, but wasn't" % unselected)
def run_scenario(self):
# Initialize topologies.
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
top_latency = Topology(
TopologySettings(seeds=seeds, monitor_class=MockMonitor,
pool_class=MockPool))
# "In latency window" is defined in the server selection
# spec as the subset of suitable_servers that falls within the
# allowable latency window.
top_suitable = Topology(
TopologySettings(seeds=seeds, local_threshold_ms=1000000,
monitor_class=MockMonitor,
pool_class=MockPool))
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
top_suitable.on_change(server_description)
top_latency.on_change(server_description)
# Create server selector.
if scenario_def["operation"] == "write":
instance = writable_server_selector
else:
# Make first letter lowercase to match read_pref's modes.
mode_string = scenario_def['read_preference']['mode']
if mode_string:
mode_string = mode_string[:1].lower() + mode_string[1:]
mode = read_preferences.read_pref_mode_from_name(mode_string)
tag_sets = None
if scenario_def['read_preference']['tag_sets'][0]:
tag_sets = scenario_def['read_preference']['tag_sets']
instance = read_preferences.make_read_preference(mode, tag_sets)
# Select servers.
if not scenario_def['suitable_servers']:
self.assertRaises(AutoReconnect, top_suitable.select_server,
instance,
server_selection_timeout=0)
return
if not scenario_def['in_latency_window']:
self.assertRaises(AutoReconnect, top_latency.select_server,
instance,
server_selection_timeout=0)
return
actual_suitable_s = top_suitable.select_servers(instance,
server_selection_timeout=0)
actual_latency_s = top_latency.select_servers(instance,
server_selection_timeout=0)
expected_suitable_servers = {}
for server in scenario_def['suitable_servers']:
server_description = make_server_description(server, hosts)
expected_suitable_servers[server['address']] = server_description
actual_suitable_servers = {}
for s in actual_suitable_s:
actual_suitable_servers["%s:%d" % (s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_suitable_servers),
len(expected_suitable_servers))
for k, actual in actual_suitable_servers.items():
expected = expected_suitable_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
expected_latency_servers = {}
for server in scenario_def['in_latency_window']:
server_description = make_server_description(server, hosts)
expected_latency_servers[server['address']] = server_description
actual_latency_servers = {}
for s in actual_latency_s:
actual_latency_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_latency_servers),
len(expected_latency_servers))
for k, actual in actual_latency_servers.items():
expected = expected_latency_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
def __init__(
self,
host=None,
port=None,
document_class=dict,
tz_aware=False,
connect=True,
**kwargs):
"""Client for a MongoDB instance, a replica set, or a set of mongoses.
The client object is thread-safe and has connection-pooling built in.
If an operation fails because of a network error,
:class:`~pymongo.errors.ConnectionFailure` is raised and the client
reconnects in the background. Application code should handle this
exception (recognizing that the operation failed) and then continue to
execute.
The `host` parameter can be a full `mongodb URI
<http://dochub.mongodb.org/core/connections>`_, in addition to
a simple hostname. It can also be a list of hostnames or
URIs. Any port specified in the host string(s) will override
the `port` parameter. If multiple mongodb URIs containing
database or auth information are passed, the last database,
username, and password present will be used. For username and
passwords reserved characters like ':', '/', '+' and '@' must be
escaped following RFC 2396.
:Parameters:
- `host` (optional): hostname or IP address of the
instance to connect to, or a mongodb URI, or a list of
hostnames / mongodb URIs. If `host` is an IPv6 literal
it must be enclosed in '[' and ']' characters following
the RFC2732 URL syntax (e.g. '[::1]' for localhost)
- `port` (optional): port number on which to connect
- `document_class` (optional): default class to use for
documents returned from queries on this client
- `tz_aware` (optional): if ``True``,
:class:`~datetime.datetime` instances returned as values
in a document by this :class:`MongoClient` will be timezone
aware (otherwise they will be naive)
- `connect` (optional): if ``True`` (the default), immediately
begin connecting to MongoDB in the background. Otherwise connect
on the first operation.
| **Other optional parameters can be passed as keyword arguments:**
- `maxPoolSize` (optional): The maximum number of connections
that the pool will open simultaneously. If this is set, operations
will block if there are `maxPoolSize` outstanding connections
from the pool. Defaults to 100. Cannot be 0.
- `socketTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait for a response after sending an
ordinary (non-monitoring) database operation before concluding that
a network error has occurred. Defaults to ``None`` (no timeout).
- `connectTimeoutMS`: (integer or None) Controls how long (in
milliseconds) the driver will wait during server monitoring when
connecting a new socket to a server before concluding the server
is unavailable. Defaults to ``20000`` (20 seconds).
- `serverSelectionTimeoutMS`: (integer) Controls how long (in
milliseconds) the driver will wait to find an available,
appropriate server to carry out a database operation; while it is
waiting, multiple server monitoring operations may be carried out,
each controlled by `connectTimeoutMS`. Defaults to ``30000`` (30
seconds).
- `waitQueueTimeoutMS`: (integer or None) How long (in milliseconds)
a thread will wait for a socket from the pool if the pool has no
free sockets. Defaults to ``None`` (no timeout).
- `waitQueueMultiple`: (integer or None) Multiplied by maxPoolSize
to give the number of threads allowed to wait for a socket at one
time. Defaults to ``None`` (no limit).
- `socketKeepAlive`: (boolean) Whether to send periodic keep-alive
packets on connected sockets. Defaults to ``False`` (do not send
keep-alive packets).
| **Write Concern options:**
| (Only set if passed. No default values.)
- `w`: (integer or string) If this is a replica set, write operations
will block until they have been replicated to the specified number
or tagged set of servers. `w=<int>` always includes the replica set
primary (e.g. w=3 means write to the primary and wait until
replicated to **two** secondaries). Passing w=0 **disables write
acknowledgement** and all other write concern options.
- `wtimeout`: (integer) Used in conjunction with `w`. Specify a value
in milliseconds to control how long to wait for write propagation
to complete. If replication does not complete in the given
timeframe, a timeout exception is raised.
- `j`: If ``True`` block until write operations have been committed
to the journal. Cannot be used in combination with `fsync`. Prior
to MongoDB 2.6 this option was ignored if the server was running
without journaling. Starting with MongoDB 2.6 write operations will
fail with an exception if this option is used when the server is
running without journaling.
- `fsync`: If ``True`` and the server is running without journaling,
blocks until the server has synced all data files to disk. If the
server is running with journaling, this acts the same as the `j`
option, blocking until write operations have been committed to the
journal. Cannot be used in combination with `j`.
| **Replica set keyword arguments for connecting with a replica set
- either directly or via a mongos:**
- `replicaSet`: (string or None) The name of the replica set to
connect to. The driver will verify that all servers it connects to
match this name. Implies that the hosts specified are a seed list
and the driver should attempt to find all members of the set.
Defaults to ``None``.
- `read_preference`: The read preference for this client. If
connecting directly to a secondary then a read preference mode
*other* than PRIMARY is required - otherwise all queries will throw
:class:`~pymongo.errors.AutoReconnect` "not master".
See :class:`~pymongo.read_preferences.ReadPreference` for all
available read preference options. Defaults to ``PRIMARY``.
| **SSL configuration:**
- `ssl`: If ``True``, create the connection to the server using SSL.
Defaults to ``False``.
- `ssl_keyfile`: The private keyfile used to identify the local
connection against mongod. If included with the ``certfile`` then
only the ``ssl_certfile`` is needed. Implies ``ssl=True``.
Defaults to ``None``.
- `ssl_certfile`: The certificate file used to identify the local
connection against mongod. Implies ``ssl=True``. Defaults to
``None``.
- `ssl_cert_reqs`: Specifies whether a certificate is required from
the other side of the connection, and whether it will be validated
if provided. It must be one of the three values ``ssl.CERT_NONE``
(certificates ignored), ``ssl.CERT_OPTIONAL``
(not required, but validated if provided), or ``ssl.CERT_REQUIRED``
(required and validated). If the value of this parameter is not
``ssl.CERT_NONE`` and a value is not provided for ``ssl_ca_certs``
PyMongo will attempt to load system provided CA certificates.
If the python version in use does not support loading system CA
certificates then the ``ssl_ca_certs`` parameter must point
to a file of CA certificates. Implies ``ssl=True``. Defaults to
``ssl.CERT_REQUIRED`` if not provided and ``ssl=True``.
- `ssl_ca_certs`: The ca_certs file contains a set of concatenated
"certification authority" certificates, which are used to validate
certificates passed from the other end of the connection.
Implies ``ssl=True``. Defaults to ``None``.
- `ssl_match_hostname`: If ``True`` (the default), and
`ssl_cert_reqs` is not ``ssl.CERT_NONE``, enables hostname
verification using the :func:`~ssl.match_hostname` function from
python's :mod:`~ssl` module. Think very carefully before setting
this to ``False`` as that could make your application vulnerable to
man-in-the-middle attacks.
.. mongodoc:: connections
.. versionchanged:: 3.0
:class:`~pymongo.mongo_client.MongoClient` is now the one and only
client class for a standalone server, mongos, or replica set.
It includes the functionality that had been split into
:class:`~pymongo.mongo_client.MongoReplicaSetClient`: it can connect
to a replica set, discover all its members, and monitor the set for
stepdowns, elections, and reconfigs.
The :class:`~pymongo.mongo_client.MongoClient` constructor no
longer blocks while connecting to the server or servers, and it no
longer raises :class:`~pymongo.errors.ConnectionFailure` if they
are unavailable, nor :class:`~pymongo.errors.ConfigurationError`
if the user's credentials are wrong. Instead, the constructor
returns immediately and launches the connection process on
background threads.
Therefore the ``alive`` method is removed since it no longer
provides meaningful information; even if the client is disconnected,
it may discover a server in time to fulfill the next operation.
In PyMongo 2.x, :class:`~pymongo.MongoClient` accepted a list of
standalone MongoDB servers and used the first it could connect to::
MongoClient(['host1.com:27017', 'host2.com:27017'])
A list of multiple standalones is no longer supported; if multiple
servers are listed they must be members of the same replica set, or
mongoses in the same sharded cluster.
The behavior for a list of mongoses is changed from "high
availability" to "load balancing". Before, the client connected to
the lowest-latency mongos in the list, and used it until a network
error prompted it to re-evaluate all mongoses' latencies and
reconnect to one of them. In PyMongo 3, the client monitors its
network latency to all the mongoses continuously, and distributes
operations evenly among those with the lowest latency. See
:ref:`mongos-load-balancing` for more information.
The ``connect`` option is added.
The ``start_request``, ``in_request``, and ``end_request`` methods
are removed, as well as the ``auto_start_request`` option.
The ``copy_database`` method is removed, see the
:doc:`copy_database examples </examples/copydb>` for alternatives.
The :meth:`MongoClient.disconnect` method is removed; it was a
synonym for :meth:`~pymongo.MongoClient.close`.
:class:`~pymongo.mongo_client.MongoClient` no longer returns an
instance of :class:`~pymongo.database.Database` for attribute names
with leading underscores. You must use dict-style lookups instead::
client['__my_database__']
Not::
client.__my_database__
"""
if host is None:
host = self.HOST
if isinstance(host, string_type):
host = [host]
if port is None:
port = self.PORT
if not isinstance(port, int):
raise TypeError("port must be an instance of int")
seeds = set()
username = None
password = None
dbase = None
opts = {}
for entity in host:
if "://" in entity:
if entity.startswith("mongodb://"):
res = uri_parser.parse_uri(entity, port, False)
seeds.update(res["nodelist"])
username = res["username"] or username
password = res["password"] or password
dbase = res["database"] or dbase
opts = res["options"]
else:
idx = entity.find("://")
raise InvalidURI("Invalid URI scheme: "
"%s" % (entity[:idx],))
else:
seeds.update(uri_parser.split_hosts(entity, port))
if not seeds:
raise ConfigurationError("need to specify at least one host")
# _pool_class, _monitor_class, and _condition_class are for deep
# customization of PyMongo, e.g. Motor.
pool_class = kwargs.pop('_pool_class', None)
monitor_class = kwargs.pop('_monitor_class', None)
condition_class = kwargs.pop('_condition_class', None)
kwargs['document_class'] = document_class
kwargs['tz_aware'] = tz_aware
opts.update(kwargs)
self.__options = options = ClientOptions(
username, password, dbase, opts)
self.__default_database_name = dbase
self.__lock = threading.Lock()
self.__cursor_manager = CursorManager(self)
self.__kill_cursors_queue = []
# Cache of existing indexes used by ensure_index ops.
self.__index_cache = {}
super(MongoClient, self).__init__(options.codec_options,
options.read_preference,
options.write_concern)
self.__all_credentials = {}
creds = options.credentials
if creds:
self._cache_credentials(creds.source, creds)
self._topology_settings = TopologySettings(
seeds=seeds,
replica_set_name=options.replica_set_name,
pool_class=pool_class,
pool_options=options.pool_options,
monitor_class=monitor_class,
condition_class=condition_class,
local_threshold_ms=options.local_threshold_ms,
server_selection_timeout=options.server_selection_timeout)
self._topology = Topology(self._topology_settings)
if connect:
self._topology.open()
def target():
client = self_ref()
if client is None:
return False # Stop the executor.
MongoClient._process_kill_cursors_queue(client)
return True
executor = periodic_executor.PeriodicExecutor(
condition_class=self._topology_settings.condition_class,
interval=common.KILL_CURSOR_FREQUENCY,
min_interval=0,
target=target)
# We strongly reference the executor and it weakly references us via
# this closure. When the client is freed, stop the executor soon.
self_ref = weakref.ref(self, executor.close)
self._kill_cursors_executor = executor
executor.open()
def run_scenario(self):
# Initialize topologies.
seeds, hosts = get_addresses(
scenario_def['topology_description']['servers'])
# "Eligible servers" is defined in the server selection spec as
# the set of servers matching both the ReadPreference's mode
# and tag sets.
top_latency = Topology(
TopologySettings(seeds=seeds,
monitor_class=MockMonitor,
pool_class=MockPool))
# "In latency window" is defined in the server selection
# spec as the subset of suitable_servers that falls within the
# allowable latency window.
top_suitable = Topology(
TopologySettings(seeds=seeds,
local_threshold_ms=1000000,
monitor_class=MockMonitor,
pool_class=MockPool))
# Update topologies with server descriptions.
for server in scenario_def['topology_description']['servers']:
server_description = make_server_description(server, hosts)
top_suitable.on_change(server_description)
top_latency.on_change(server_description)
# Create server selector.
if scenario_def["operation"] == "write":
instance = writable_server_selector
else:
# Make first letter lowercase to match read_pref's modes.
mode_string = scenario_def['read_preference']['mode']
if mode_string:
mode_string = mode_string[:1].lower() + mode_string[1:]
mode = read_preferences.read_pref_mode_from_name(mode_string)
tag_sets = None
if scenario_def['read_preference']['tag_sets'][0]:
tag_sets = scenario_def['read_preference']['tag_sets']
instance = read_preferences.make_read_preference(mode, tag_sets)
# Select servers.
if not scenario_def['suitable_servers']:
self.assertRaises(AutoReconnect,
top_suitable.select_server,
instance,
server_selection_timeout=0)
return
if not scenario_def['in_latency_window']:
self.assertRaises(AutoReconnect,
top_latency.select_server,
instance,
server_selection_timeout=0)
return
actual_suitable_s = top_suitable.select_servers(
instance, server_selection_timeout=0)
actual_latency_s = top_latency.select_servers(
instance, server_selection_timeout=0)
expected_suitable_servers = {}
for server in scenario_def['suitable_servers']:
server_description = make_server_description(server, hosts)
expected_suitable_servers[server['address']] = server_description
actual_suitable_servers = {}
for s in actual_suitable_s:
actual_suitable_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_suitable_servers),
len(expected_suitable_servers))
for k, actual in actual_suitable_servers.items():
expected = expected_suitable_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
expected_latency_servers = {}
for server in scenario_def['in_latency_window']:
server_description = make_server_description(server, hosts)
expected_latency_servers[server['address']] = server_description
actual_latency_servers = {}
for s in actual_latency_s:
actual_latency_servers["%s:%d" %
(s.description.address[0],
s.description.address[1])] = s.description
self.assertEqual(len(actual_latency_servers),
len(expected_latency_servers))
for k, actual in actual_latency_servers.items():
expected = expected_latency_servers[k]
self.assertEqual(expected.address, actual.address)
self.assertEqual(expected.server_type, actual.server_type)
self.assertEqual(expected.round_trip_time, actual.round_trip_time)
self.assertEqual(expected.tags, actual.tags)
self.assertEqual(expected.all_hosts, actual.all_hosts)
