13 Distributed Erlang
13.1 Distributed Erlang System
A distributed Erlang system consists of a number of Erlang runtime systems communicating with each other. Each such runtime system is called a node. Message passing between processes at different nodes, as well as links and monitors, are transparent when pids are used. Registered names, however, are local to each node. This means that the node must be specified as well when sending messages, and so on, using registered names.
The distribution mechanism is implemented using TCP/IP sockets. How to implement an alternative carrier is described in the ERTS User's Guide.
Starting a distributed node without also specifying -proto_dist inet_tls will expose the node to attacks that may give the attacker complete access to the node and in extension the cluster. When using un-secure distributed nodes, make sure that the network is configured to keep potential attackers out. See the Using SSL for Erlang Distribution User's Guide for details on how to setup a secure distributed node.
13.2 Nodes
A node is an executing Erlang runtime system that has been given a name, using the command-line flag -name (long names) or -sname (short names).
The format of the node name is an atom name@host. name is the name given by the user. host is the full host name if long names are used, or the first part of the host name if short names are used. node() returns the name of the node.
Example:
% erl -name dilbert ([email protected])1> node(). '[email protected]' % erl -sname dilbert (dilbert@uab)1> node(). dilbert@uab
A node with a long node name cannot communicate with a node with a short node name.
13.3 Node Connections
The nodes in a distributed Erlang system are loosely connected. The first time the name of another node is used, for example, if spawn(Node,M,F,A) or net_adm:ping(Node) is called, a connection attempt to that node is made.
Connections are by default transitive. If a node A connects to node B, and node B has a connection to node C, then node A also tries to connect to node C. This feature can be turned off by using the command-line flag -connect_all false, see the erl(1) manual page in ERTS.
If a node goes down, all connections to that node are removed. Calling erlang:disconnect_node(Node) forces disconnection of a node.
The list of (visible) nodes currently connected to is returned by nodes().
13.4 epmd
The Erlang Port Mapper Daemon epmd is automatically started at every host where an Erlang node is started. It is responsible for mapping the symbolic node names to machine addresses. See the epmd(1) manual page in ERTS.
13.5 Hidden Nodes
In a distributed Erlang system, it is sometimes useful to connect to a node without also connecting to all other nodes. An example is some kind of O&M functionality used to inspect the status of a system, without disturbing it. For this purpose, a hidden node can be used.
A hidden node is a node started with the command-line flag -hidden. Connections between hidden nodes and other nodes are not transitive, they must be set up explicitly. Also, hidden nodes does not show up in the list of nodes returned by nodes(). Instead, nodes(hidden) or nodes(connected) must be used. This means, for example, that the hidden node is not added to the set of nodes that global is keeping track of.
This feature was added in Erlang 5.0/OTP R7.
13.6 C Nodes
A C node is a C program written to act as a hidden node in a distributed Erlang system. The library Erl_Interface contains functions for this purpose. For more information about C nodes, see the Erl_Interface application and Interoperability Tutorial..
13.7 Security
Authentication determines which nodes are allowed to communicate with each other. In a network of different Erlang nodes, it is built into the system at the lowest possible level. Each node has its own magic cookie, which is an Erlang atom.
When a node tries to connect to another node, the magic cookies are compared. If they do not match, the connected node rejects the connection.
At start-up, a node has a random atom assigned as its magic cookie and the cookie of other nodes is assumed to be nocookie. The first action of the Erlang network authentication server (auth) is then to read a file named $HOME/.erlang.cookie. If the file does not exist, it is created. The UNIX permissions mode of the file is set to octal 400 (read-only by user) and its contents are a random string. An atom Cookie is created from the contents of the file and the cookie of the local node is set to this using erlang:set_cookie(node(), Cookie). This also makes the local node assume that all other nodes have the same cookie Cookie.
Thus, groups of users with identical cookie files get Erlang nodes that can communicate freely and without interference from the magic cookie system. Users who want to run nodes on separate file systems must make certain that their cookie files are identical on the different file systems.
For a node Node1 with magic cookie Cookie to be able to connect to, or accept a connection from, another node Node2 with a different cookie DiffCookie, the function erlang:set_cookie(Node2, DiffCookie) must first be called at Node1. Distributed systems with multiple user IDs can be handled in this way.
The default when a connection is established between two nodes, is to immediately connect all other visible nodes as well. This way, there is always a fully connected network. If there are nodes with different cookies, this method can be inappropriate and the command-line flag -connect_all false must be set, see the erl(1) manual page in ERTS.
The magic cookie of the local node is retrieved by calling erlang:get_cookie().
13.8 Distribution BIFs
Some useful BIFs for distributed programming (for more information, see the erlang(3) manual page in ERTS:
| BIF | Description |
| erlang:disconnect_node(Node) | Forces the disconnection of a node. |
| erlang:get_cookie() | Returns the magic cookie of the current node. |
| is_alive() | Returns true if the runtime system is a node and can connect to other nodes, false otherwise. |
| monitor_node(Node, true|false) | Monitors the status of Node. A message{nodedown, Node} is received if the connection to it is lost. |
| node() | Returns the name of the current node. Allowed in guards. |
| node(Arg) | Returns the node where Arg, a pid, reference, or port, is located. |
| nodes() | Returns a list of all visible nodes this node is connected to. |
| nodes(Arg) | Depending on Arg, this function can return a list not only of visible nodes, but also hidden nodes and previously known nodes, and so on. |
| erlang:set_cookie(Node, Cookie) | Sets the magic cookie used when connecting to Node. If Node is the current node, Cookie is used when connecting to all new nodes. |
| spawn[_link|_opt](Node, Fun) | Creates a process at a remote node. |
| spawn[_link|opt](Node, Module, FunctionName, Args) | Creates a process at a remote node. |
Table 13.1: Distribution BIFs
13.9 Distribution Command-Line Flags
Examples of command-line flags used for distributed programming (for more information, see the erl(1) manual page in ERTS:
| Command-Line Flag | Description |
| -connect_all false | Only explicit connection set-ups are used. |
| -hidden | Makes a node into a hidden node. |
| -name Name | Makes a runtime system into a node, using long node names. |
| -setcookie Cookie | Same as calling erlang:set_cookie(node(), Cookie). |
| -sname Name | Makes a runtime system into a node, using short node names. |
Table 13.2: Distribution Command-Line Flags
13.10 Distribution Modules
Examples of modules useful for distributed programming:
In the Kernel application:
| Module | Description |
| global | A global name registration facility. |
| global_group | Grouping nodes to global name registration groups. |
| net_adm | Various Erlang net administration routines. |
| net_kernel | Erlang networking kernel. |
Table 13.3: Kernel Modules Useful For Distribution.
In the STDLIB application:
| Module | Description |
| slave | Start and control of slave nodes. |
Table 13.4: STDLIB Modules Useful For Distribution.