Config¶

Hazelcast Client Configuration module contains configuration classes and various constants required to create a ClientConfig.

DEFAULT_GROUP_NAME = 'dev'¶: Default group name of the connected Hazelcast cluster

DEFAULT_GROUP_PASSWORD = 'dev-pass'¶: Default password of connected Hazelcast cluster

PROPERTY_HEARTBEAT_INTERVAL = 'hazelcast.client.heartbeat.interval'¶: Configuration property for heartbeat interval in milliseconds. Client will send heartbeat to server by this value of interval unless other requests send to server

PROPERTY_HEARTBEAT_TIMEOUT = 'hazelcast.client.heartbeat.timeout'¶: Configuration property for heartbeat timeout in milliseconds. If client cannot see any activity on a connection for this timeout value it will shutdown the connection

INTEGER_TYPE¶

Integer type options that can be used by serialization service.

VAR : variable size integer (this option can be problematic on static type clients like java or .NET)
BYTE: Python int will be interpreted as a single byte int
SHORT: Python int will be interpreted as a double byte int
INT: Python int will be interpreted as a four byte int
LONG: Python int will be interpreted as an eight byte int
BIG_INT: Python int will be interpreted as Java BigInteger. This option can handle python long values with “bit_length > 64”

alias of Enum

EVICTION_POLICY¶

Near Cache eviction policy options

NONE : No evcition
LRU : Least Recently Used items will be evicted
LFU : Least frequently Used items will be evicted
RANDOM : Items will be evicted randomly

alias of Enum

IN_MEMORY_FORMAT¶

Near Cache in memory format of the values.

BINARY : Binary format, hazelcast serializated bytearray format
OBJECT : The actual objects used

alias of Enum

class ClientConfig¶

Bases: object

The root configuration for hazelcast python client.

>>> client_config = ClientConfig()
>>> client = HazelcastClient(client_config)

group_config = None¶: The group configuration

network_config = None¶: The network configuration for addresses to connect, smart-routing, socket-options...

load_balancer = None¶: Custom load balancer used to distribute the operations to multiple Endpoints.

membership_listeners = None¶: Membership listeners, an array of tuple (member_added, member_removed, fire_for_existing)

lifecycle_listeners = None¶: Lifecycle Listeners, an array of Functions of f(state)

near_cache_configs = None¶: Near Cache configuration which maps “map-name : NearCacheConfig

serialization_config = None¶: Hazelcast serialization configuration

add_membership_listener(member_added=None, member_removed=None, fire_for_existing=False)¶

Helper method for adding membership listeners

Parameters:	member_added – (Function), Function to be called when a member is added, in the form of f(member)

(optional). :param member_removed: (Function), Function to be called when a member is removed, in the form of f(member) (optional). :param fire_for_existing: if True, already existing members will fire member_added event (optional). :return: self for cascading configuration

add_lifecycle_listener(lifecycle_state_changed=None)¶

Helper method for adding lifecycle listeners.

Parameters:	lifecycle_state_changed – (Function), Function to be called when lifecycle state is changed (optional).

In the form of f(state). :return: self for cascading configuration

get_property_or_default(key, default)¶

Client property accessor with fallback to default value.

Parameters:	key – (Object), property key to access. default – (Object), the default value for fallback.
Returns:	(Object), property value if it exist or the default value otherwise.

add_near_cache_config(near_cache_config)¶

Helper method to add a new NearCacheConfig.

Parameters:	near_cache_config – (NearCacheConfig), the near_cache config to add.
Returns:	self for cascading configuration.

class GroupConfig¶

Bases: object

The Group Configuration is the container class for name and password of the cluster.

name = None¶: The group name of the cluster

password = None¶: The password of the cluster

class ClientNetworkConfig¶

Bases: object

Network related configuration parameters.

addresses = None¶: The candidate address list that client will use to establish initial connection

connection_attempt_limit = None¶: While client is trying to connect initially to one of the members in the addressList, all might be not available. Instead of giving up, throwing Error and stopping client, it will attempt to retry as much as defined by this parameter.

connection_attempt_period = None¶: Period for the next attempt to find a member to connect

connection_timeout = None¶: Socket connection timeout is a float, giving in seconds, or None. Setting a timeout of None disables the timeout feature and is equivalent to block the socket until it connects. Setting a timeout of zero is the same as disables blocking on connect.

socket_options = None¶

Array of Unix socket options.

Example usage:

>>> import socket
>>> client_network_config.socket_options.append(SocketOption(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1))
>>> client_network_config.socket_options.append(SocketOption(socket.SOL_SOCKET, socket.SO_SNDBUF, 32768))
>>> client_network_config.socket_options.append(SocketOption(socket.SOL_SOCKET, socket.SO_RCVBUF, 32768))

Please see the Unix manual for level and option. Level and option constant are in python std lib socket module

redo_operation = None¶: If true, client will redo the operations that were executing on the server and client lost the connection. This can be because of network, or simply because the member died. However it is not clear whether the application is performed or not. For idempotent operations this is harmless, but for non idempotent ones retrying can cause to undesirable effects. Note that the redo can perform on any member.

smart_routing = None¶: If true, client will route the key based operations to owner of the key at the best effort. Note that it uses a cached value of partition count and doesn’t guarantee that the operation will always be executed on the owner. The cached table is updated every 10 seconds.

class SocketOption(level, option, value)¶

Bases: object

Advanced configuration for fine-tune the TCP options. A Socket option represent the unix socket option, that will be passed to python socket.setoption(level,`option, value)` See the Unix manual for level and option.

level = None¶: Option level. See the Unix manual for detail.

option = None¶: The actual socket option. The actual socket option.

value = None¶: Socket option value. The value argument can either be an integer or a string

class SerializationConfig¶

Bases: object

Hazelcast Serialization Service configuration options can be set from this class.

portable_version = None¶: Portable version will be used to differentiate two versions of the same class that have changes on the class, like adding/removing a field or changing a type of a field.

data_serializable_factories = None¶

Dictionary of factory-id and corresponding IdentifiedDataserializable factories. A Factory is a simple dictionary with entries of class-id : class-constructor-function pairs.

Example:

>>> my_factory = {MyPersonClass.CLASS_ID : MyPersonClass, MyAddressClass.CLASS_ID : MyAddressClass}
>>> serialization_config.data_serializable_factories[FACTORY_ID] = my_factory

portable_factories = None¶

Dictionary of factory-id and corresponding portable factories. A Factory is a simple dictionary with entries of class-id : class-constructor-function pairs.

Example:

>>> portable_factory = {PortableClass_0.CLASS_ID : PortableClass_0, PortableClass_1.CLASS_ID : PortableClass_1}
>>> serialization_config.portable_factories[FACTORY_ID] = portable_factory

class_definitions = None¶: Set of all Portable class definitions.

check_class_def_errors = None¶: Configured Portable Class definitions should be validated for errors or not.

is_big_endian = None¶: Hazelcast Serialization is big endian or not.

default_integer_type = None¶: Python has variable length int/long type. In order to match this with static fixed length Java server, this option defines the length of the int/long. One of the values of INTEGER_TYPE can be assigned. Please see INTEGER_TYPE documentation for details of the options.

add_data_serializable_factory(factory_id, factory)¶

Helper method for adding IdentifiedDataSerializable factory. example:

>>> my_factory = {MyPersonClass.CLASS_ID : MyPersonClass, MyAddressClass.CLASS_ID : MyAddressClass}
>>> serialization_config.add_data_serializable_factory(factory_id, my_factory)

Parameters:	factory_id – (int), factory-id to register. factory – (Dictionary), the factory dictionary of class-id:class-constructor-function pairs.

add_portable_factory(factory_id, factory)¶

Helper method for adding Portable factory. example:

>>> portable_factory = {PortableClass_0.CLASS_ID : PortableClass_0, PortableClass_1.CLASS_ID : PortableClass_1}
>>> serialization_config.portable_factories[FACTORY_ID] = portable_factory

Parameters:	factory_id – (int), factory-id to register. factory – (Dictionary), the factory dictionary of class-id:class-constructor-function pairs.

set_custom_serializer(_type, serializer)¶

Assign a serializer for the type.

Parameters:	_type – (Type), the target type of the serializer serializer – (Serializer), Custom Serializer constructor function

custom_serializers¶

All custom serializers.

Returns:	(Dictionary), dictionary of type-custom serializer pairs.

global_serializer¶

The Global serializer property for serialization service. The assigned value should be a class constructor function. It handles every object if no other serializer found.

Global serializers should extend hazelcast.serializer.api.StreamSerializer

class NearCacheConfig(name)¶

Bases: object

Map Near cache configuration for a specific map by name.

invalidate_on_change = None¶: Should a value is invalidated and removed in case of any map data updating operations such as replace, remove etc.

name¶: Name of the map that this near cache belong. Cannot be None.

in_memory_format¶: Internal representation of the stored data in near cache.

time_to_live_seconds¶: The maximum number of seconds for each entry to stay in the near cache.

max_idle_seconds¶: Maximum number of seconds each entry can stay in the near cache as untouched (not-read).

eviction_policy¶: The eviction policy for the near cache

eviction_max_size¶: The limit for number of entries until the eviction start.

eviction_sampling_count¶: The entry count of the samples for the internal eviction sampling algorithm taking samples in each operation.

eviction_sampling_pool_size¶: The size of the internal eviction sampling algorithm has a pool of best candidates for eviction.