path: root/lib/kernel/doc/src/application.xml

<?xml version="1.0" encoding="latin1" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">

<erlref>
  <header>
    <copyright>
      <year>1996</year><year>2009</year>
      <holder>Ericsson AB. All Rights Reserved.</holder>
    </copyright>
    <legalnotice>
      The contents of this file are subject to the Erlang Public License,
      Version 1.1, (the "License"); you may not use this file except in
      compliance with the License. You should have received a copy of the
      Erlang Public License along with this software. If not, it can be
      retrieved online at http://www.erlang.org/.
    
      Software distributed under the License is distributed on an "AS IS"
      basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
      the License for the specific language governing rights and limitations
      under the License.
    
    </legalnotice>

    <title>application</title>
    <prepared></prepared>
    <docno></docno>
    <date></date>
    <rev></rev>
  </header>
  <module>application</module>
  <modulesummary>Generic OTP application functions</modulesummary>
  <description>
    <p>In OTP, <em>application</em> denotes a component implementing
      some specific functionality, that can be started and stopped as a
      unit, and which can be re-used in other systems as well. This
      module interfaces the <em>application controller</em>, a process
      started at every Erlang runtime system, and contains functions
      for controlling applications (for example starting and stopping
      applications), and functions to access information about
      applications (for example configuration parameters).</p>
    <p>An application is defined by an <em>application specification</em>. The specification is normally located in an
      <em>application resource file</em> called <c>Application.app</c>,
      where <c>Application</c> is the name of the application. Refer to
      <seealso marker="app">app(4)</seealso> for more information about
      the application specification.</p>
    <p>This module can also be viewed as a behaviour for an application
      implemented according to the OTP design principles as a
      supervision tree. The definition of how to start and stop
      the tree should be located in an <em>application callback module</em> exporting a pre-defined set of functions.</p>
    <p>Refer to <seealso marker="doc/design_principles:des_princ">OTP Design Principles</seealso> for more information about
      applications and behaviours.</p>
  </description>
  <funcs>
    <func>
      <name>get_all_env() -> Env</name>
      <name>get_all_env(Application) -> Env</name>
      <fsummary>Get the configuration parameters for an application</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Env = [{Par,Val}]</v>
        <v>&nbsp;Par = atom()</v>
        <v>&nbsp;Val = term()</v>
      </type>
      <desc>
        <p>Returns the configuration parameters and their values for
          <c>Application</c>. If the argument is omitted, it defaults to
          the application of the calling process.</p>
        <p>If the specified application is not loaded, or if the process
          executing the call does not belong to any application,
          the function returns <c>[]</c>.</p>
      </desc>
    </func>
    <func>
      <name>get_all_key() -> {ok, Keys} | []</name>
      <name>get_all_key(Application) -> {ok, Keys} | undefined </name>
      <fsummary>Get the application specification keys</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Keys = [{Key,Val}]</v>
        <v>&nbsp;Key = atom()</v>
        <v>&nbsp;Val = term()</v>
      </type>
      <desc>
        <p>Returns the application specification keys and their values
          for <c>Application</c>. If the argument is omitted, it
          defaults to the application of the calling process.</p>
        <p>If the specified application is not loaded, the function
          returns <c>undefined</c>. If the process executing the call
          does not belong to any application, the function returns
          <c>[]</c>.</p>
      </desc>
    </func>
    <func>
      <name>get_application() -> {ok, Application} | undefined</name>
      <name>get_application(Pid | Module) -> {ok, Application} | undefined</name>
      <fsummary>Get the name of an application containing a certain process or module</fsummary>
      <type>
        <v>Pid = pid()</v>
        <v>Module = atom()</v>
        <v>Application = atom()</v>
      </type>
      <desc>
        <p>Returns the name of the application to which the process
          <c>Pid</c> or the module <c>Module</c> belongs. Providing no
          argument is the same as calling
          <c>get_application(self())</c>.</p>
        <p>If the specified process does not belong to any application,
          or if the specified process or module does not exist,
          the function returns <c>undefined</c>.</p>
      </desc>
    </func>
    <func>
      <name>get_env(Par) -> {ok, Val} | undefined</name>
      <name>get_env(Application, Par) -> {ok, Val} | undefined</name>
      <fsummary>Get the value of a configuration parameter</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Par = atom()</v>
        <v>Val = term()</v>
      </type>
      <desc>
        <p>Returns the value of the configuration parameter <c>Par</c>
          for <c>Application</c>. If the application argument is
          omitted, it defaults to the application of the calling
          process.</p>
        <p>If the specified application is not loaded, or
          the configuration parameter does not exist, or if the process
          executing the call does not belong to any application,
          the function returns <c>undefined</c>.</p>
      </desc>
    </func>
    <func>
      <name>get_key(Key) -> {ok, Val} | undefined</name>
      <name>get_key(Application, Key) -> {ok, Val} | undefined</name>
      <fsummary>Get the value of an application specification key</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Key = atom()</v>
        <v>Val = term()</v>
      </type>
      <desc>
        <p>Returns the value of the application specification key
          <c>Key</c> for <c>Application</c>. If the application
          argument is omitted, it defaults to the application of
          the calling process.</p>
        <p>If the specified application is not loaded, or
          the specification key does not exist, or if the process
          executing the call does not belong to any application,
          the function returns <c>undefined</c>.</p>
      </desc>
    </func>
    <func>
      <name>load(AppDescr) -> ok | {error, Reason}</name>
      <name>load(AppDescr, Distributed) -> ok | {error, Reason}</name>
      <fsummary>Load an application</fsummary>
      <type>
        <v>AppDescr = Application | AppSpec</v>
        <v>&nbsp;Application = atom()</v>
        <v>&nbsp;AppSpec = {application,Application,AppSpecKeys}</v>
        <v>&nbsp;&nbsp;AppSpec = [{Key,Val}]</v>
        <v>&nbsp;&nbsp;&nbsp;Key = atom()</v>
        <v>&nbsp;&nbsp;&nbsp;Val = term()</v>
        <v>Distributed = {Application,Nodes} | {Application,Time,Nodes} | default</v>
        <v>&nbsp;Nodes = [node() | {node(),..,node()}]</v>
        <v>&nbsp;Time = integer() > 0</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Loads the application specification for an application into
          the application controller. It will also load the application
          specifications for any included applications. Note that
          the function does not load the actual Erlang object code.</p>
        <p>The application can be given by its name <c>Application</c>.
          In this case the application controller will search the code
          path for the application resource file <c>Application.app</c>
          and load the specification it contains.</p>
        <p>The application specification can also be given directly as a
          tuple <c>AppSpec</c>. This tuple should have the format and
          contents as described in <c>app(4)</c>.</p>
        <p>If <c>Distributed == {Application,[Time,]Nodes}</c>,
          the application will be distributed. The argument overrides
          the value for the application in the Kernel configuration
          parameter <c>distributed</c>. <c>Application</c> must be
          the name of the application (same as in the first argument).
          If a node crashes and <c>Time</c> has been specified, then
          the application controller will wait for <c>Time</c>
          milliseconds before attempting to restart the application on
          another node. If <c>Time</c> is not specified, it will
          default to 0 and the application will be restarted
          immediately.</p>
        <p><c>Nodes</c> is a list of node names where the application
          may run, in priority from left to right. Node names can be
          grouped using tuples to indicate that they have the same
          priority. Example:</p>
        <code type="none">
Nodes = [cp1@cave, {cp2@cave, cp3@cave}]</code>
        <p>This means that the application should preferably be started
          at <c>cp1@cave</c>. If <c>cp1@cave</c> is down,
          the application should be started at either <c>cp2@cave</c>
          or <c>cp3@cave</c>.</p>
        <p>If <c>Distributed == default</c>, the value for
          the application in the Kernel configuration parameter
          <c>distributed</c> will be used.</p>
      </desc>
    </func>
    <func>
      <name>loaded_applications() -> [{Application, Description, Vsn}]</name>
      <fsummary>Get the currently loaded applications</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Description = string()</v>
        <v>Vsn = string()</v>
      </type>
      <desc>
        <p>Returns a list with information about the applications which
          have been loaded using <c>load/1,2</c>, also included
          applications. <c>Application</c> is the application name.
          <c>Description</c> and <c>Vsn</c> are the values of its
          <c>description</c> and <c>vsn</c> application specification
          keys, respectively.</p>
      </desc>
    </func>
    <func>
      <name>permit(Application, Bool) -> ok | {error, Reason}</name>
      <fsummary>Change an application's permission to run on a node.</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Bool = bool()</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Changes the permission for <c>Application</c> to run at
          the current node. The application must have been loaded using
          <c>load/1,2</c> for the function to have effect.</p>
        <p>If the permission of a loaded, but not started, application
          is set to <c>false</c>, <c>start</c> will return <c>ok</c> but
          the application will not be started until the permission is
          set to <c>true</c>.</p>
        <p>If the permission of a running application is set to
          <c>false</c>, the application will be stopped. If
          the permission later is set to <c>true</c>, it will be
          restarted.</p>
        <p>If the application is distributed, setting the permission to
          <c>false</c> means that the application will be started at, or
          moved to, another node according to how its distribution is
          configured (see <c>load/2</c> above).</p>
        <p>The function does not return until the application is
          started, stopped or successfully moved to another node.
          However, in some cases where permission is set to <c>true</c>
          the function may return <c>ok</c> even though the application
          itself has not started. This is true when an application
          cannot start because it has dependencies to other
          applications which have not yet been started. When they have
          been started, <c>Application</c> will be started as well.</p>
        <p>By default, all applications are loaded with permission
          <c>true</c> on all nodes. The permission is configurable by
          using the Kernel configuration parameter <c>permissions</c>.</p>
      </desc>
    </func>
    <func>
      <name>set_env(Application, Par, Val) -> ok</name>
      <name>set_env(Application, Par, Val, Timeout) -> ok</name>
      <fsummary>Set the value of a configuration parameter</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Par = atom()</v>
        <v>Val = term()</v>
        <v>Timeout = int() | infinity</v>
      </type>
      <desc>
        <p>Sets the value of the configuration parameter <c>Par</c> for
          <c>Application</c>.</p>
        <p><c>set_env/3</c> uses the standard <c>gen_server</c> timeout
          value (5000 ms). A <c>Timeout</c> argument can be provided
          if another timeout value is useful, for example, in situations
          where the application controller is heavily loaded.</p>
        <warning>
          <p>Use this function only if you know what you are doing, 
            that is, on your own applications. It is very application
            and configuration parameter dependent when and how often
            the value is read by the application, and careless use 
            of this function may put the application in a 
            weird, inconsistent, and malfunctioning state. </p>
        </warning>
      </desc>
    </func>
    <func>
      <name>start(Application) -> ok | {error, Reason}</name>
      <name>start(Application, Type) -> ok | {error, Reason}</name>
      <fsummary>Load and start an application</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Type = permanent | transient | temporary</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Starts <c>Application</c>. If it is not loaded,
          the application controller will first load it using
          <c>load/1</c>. It will make sure any included applications
          are loaded, but will not start them. That is assumed to be
          taken care of in the code for <c>Application</c>.</p>
        <p>The application controller checks the value of
          the application specification key <c>applications</c>, to
          ensure that all applications that should be started before
          this application are running. If not,
          <c>{error,{not_started,App}}</c> is returned, where <c>App</c>
          is the name of the missing application.</p>
        <p>The application controller then creates an <em>application master</em> for the application. The application master is
          the group leader of all the processes in the application.
          The application master starts the application by calling
          the application callback function <c>Module:start/2</c> as
          defined by the application specification key <c>mod</c>.</p>
        <p>The <c>Type</c> argument specifies the type of
          the application. If omitted, it defaults to <c>temporary</c>.</p>
        <list type="bulleted">
          <item>If a permanent application terminates, all other
           applications and the entire Erlang node are also terminated.</item>
          <item>If a transient application terminates with <c>Reason == normal</c>, this is reported but no other applications are
           terminated. If a transient application terminates
           abnormally, all other applications and the entire Erlang
           node are also terminated.</item>
          <item>If a temporary application terminates, this is reported
           but no other applications are terminated.</item>
        </list>
        <p>Note that it is always possible to stop an application
          explicitly by calling <c>stop/1</c>. Regardless of the type of
          the application, no other applications will be affected.</p>
        <p>Note also that the transient type is of little practical use,
          since when a supervision tree terminates, the reason is set to
          <c>shutdown</c>, not <c>normal</c>.</p>
      </desc>
    </func>
    <func>
      <name>start_type() -> StartType | local | undefined</name>
      <fsummary>Get the start type of an ongoing application startup.</fsummary>
      <type>
        <v>StartType = normal | {takeover,Node} | {failover,Node}</v>
        <v>&nbsp;Node = node()</v>
      </type>
      <desc>
        <p>This function is intended to be called by a process belonging
          to an application, when the application is being started, to
          determine the start type which is either <c>StartType</c> or
          <c>local</c>.</p>
        <p>See <c>Module:start/2</c> for a description of
          <c>StartType</c>.</p>
        <p><c>local</c> is returned if only parts of the application is
          being restarted (by a supervisor), or if the function is
          called outside a startup.</p>
        <p>If the process executing the call does not belong to any
          application, the function returns <c>undefined</c>.</p>
      </desc>
    </func>
    <func>
      <name>stop(Application) -> ok | {error, Reason}</name>
      <fsummary>Stop an application</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Stops <c>Application</c>. The application master calls
          <c>Module:prep_stop/1</c>, if such a function is defined, and
          then tells the top supervisor of the application to shutdown
          (see <c>supervisor(3)</c>). This means that the entire
          supervision tree, including included applications, is
          terminated in reversed start order. After the shutdown,
          the application master calls <c>Module:stop/1</c>.
          <c>Module</c> is the callback module as defined by
          the application specification key <c>mod</c>.</p>
        <p>Last, the application master itself terminates. Note that all
          processes with the application master as group leader, i.e.
          processes spawned from a process belonging to the application,
          thus are terminated as well.</p>
        <p>When stopped, the application is still loaded.</p>
        <p>In order to stop a distributed application, <c>stop/1</c>
          has to be called on all nodes where it can execute (that is,
          on all nodes where it has been started). The call to
          <c>stop/1</c> on the node where the application currently
          executes will stop its execution. The application will not be
          moved between nodes due to <c>stop/1</c> being called on
          the node where the application currently executes before
          <c>stop/1</c> is called on the other nodes.</p>
      </desc>
    </func>
    <func>
      <name>takeover(Application, Type) -> ok | {error, Reason}</name>
      <fsummary>Take over a distributed application</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Type = permanent | transient | temporary</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Performs a takeover of the distributed application
          <c>Application</c>, which executes at another node
          <c>Node</c>. At the current node, the application is
          restarted by calling
          <c>Module:start({takeover,Node},StartArgs)</c>. <c>Module</c>
          and <c>StartArgs</c> are retrieved from the loaded application
          specification. The application at the other node is not
          stopped until the startup is completed, i.e. when
          <c>Module:start/2</c> and any calls to
          <c>Module:start_phase/3</c> have returned.</p>
        <p>Thus two instances of the application will run simultaneously
          during the takeover, which makes it possible to transfer data
          from the old to the new instance. If this is not acceptable
          behavior, parts of the old instance may be shut down when
          the new instance is started. Note that the application may
          not be stopped entirely however, at least the top supervisor
          must remain alive.</p>
        <p>See <c>start/1,2</c> for a description of <c>Type</c>.</p>
      </desc>
    </func>
    <func>
      <name>unload(Application) -> ok | {error, Reason}</name>
      <fsummary>Unload an application</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Reason = term()</v>
      </type>
      <desc>
        <p>Unloads the application specification for <c>Application</c>
          from the application controller. It will also unload
          the application specifications for any included applications.
          Note that the function does not purge the actual Erlang
          object code.</p>
      </desc>
    </func>
    <func>
      <name>unset_env(Application, Par) -> ok</name>
      <name>unset_env(Application, Par, Timeout) -> ok</name>
      <fsummary>Unset the value of a configuration parameter</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Par = atom()</v>
        <v>Timeout = int() | infinity</v>
      </type>
      <desc>
        <p>Removes the configuration parameter <c>Par</c> and its value
          for <c>Application</c>.</p>
        <p><c>unset_env/2</c> uses the standard <c>gen_server</c>
          timeout value (5000 ms). A <c>Timeout</c> argument can be
          provided if another timeout value is useful, for example, in
          situations where the application controller is heavily loaded.</p>
        <warning>
          <p>Use this function only if you know what you are doing, 
            that is, on your own applications. It is very application
            and configuration parameter dependent when and how often
            the value is read by the application, and careless use 
            of this function may put the application in a 
            weird, inconsistent, and malfunctioning state. </p>
        </warning>
      </desc>
    </func>
    <func>
      <name>which_applications() -> [{Application, Description, Vsn}]</name>
      <name>which_applications(Timeout) -> [{Application, Description, Vsn}]</name>
      <fsummary>Get the currently running applications</fsummary>
      <type>
        <v>Application = atom()</v>
        <v>Description = string()</v>
        <v>Vsn = string()</v>
        <v>Timeout = int() | infinity</v>
      </type>
      <desc>
        <p>Returns a list with information about the applications which
          are currently running. <c>Application</c> is the application
          name. <c>Description</c> and <c>Vsn</c> are the values of its
          <c>description</c> and <c>vsn</c> application specification
          keys, respectively.</p>
        <p><c>which_applications/0</c> uses the standard
          <c>gen_server</c> timeout value (5000 ms). A <c>Timeout</c>
          argument can be provided if another timeout value is useful,
          for example, in situations where the application controller
          is heavily loaded.</p>
      </desc>
    </func>
  </funcs>

  <section>
    <title>CALLBACK MODULE</title>
    <p>The following functions should be exported from an
      <c>application</c> callback module.</p>
  </section>
  <funcs>
    <func>
      <name>Module:start(StartType, StartArgs) -> {ok, Pid} | {ok, Pid, State} | {error, Reason}</name>
      <fsummary>Start an application</fsummary>
      <type>
        <v>StartType = normal | {takeover,Node} | {failover,Node}</v>
        <v>&nbsp;Node = node()</v>
        <v>StartArgs = term()</v>
        <v>Pid = pid()</v>
        <v>State = term()</v>
      </type>
      <desc>
        <p>This function is called whenever an application is started
          using <c>application:start/1,2</c>, and should start
          the processes of the application. If the application is
          structured according to the OTP design principles as a
          supervision tree, this means starting the top supervisor of
          the tree.</p>
        <p><c>StartType</c> defines the type of start:</p>
        <list type="bulleted">
          <item><c>normal</c> if it's a normal startup.</item>
          <item><c>normal</c> also if the application is distributed and
           started at the current node due to a failover from another
           node, and the application specification key <c>start_phases == undefined</c>.</item>
          <item><c>{takeover,Node}</c> if the application is
           distributed and started at the current node due to a
           takeover from <c>Node</c>, either because
          <c>application:takeover/2</c> has been called or because
           the current node has higher priority than <c>Node</c>.</item>
          <item><c>{failover,Node}</c> if the application is
           distributed and started at the current node due to a
           failover from <c>Node</c>, and the application
           specification key <c>start_phases /= undefined</c>.</item>
        </list>
        <p><c>StartArgs</c> is the <c>StartArgs</c> argument defined by
          the application specification key <c>mod</c>.</p>
        <p>The function should return <c>{ok,Pid}</c> or
          <c>{ok,Pid,State}</c> where <c>Pid</c> is the pid of the top
          supervisor and <c>State</c> is any term. If omitted,
          <c>State</c> defaults to <c>[]</c>. If later the application
          is stopped, <c>State</c> is passed to
          <c>Module:prep_stop/1</c>.</p>
      </desc>
    </func>
    <func>
      <name>Module:start_phase(Phase, StartType, PhaseArgs) -> ok | {error, Reason}</name>
      <fsummary>Extended start of an application</fsummary>
      <type>
        <v>Phase = atom()</v>
        <v>StartType = normal | {takeover,Node} | {failover,Node}</v>
        <v>&nbsp;Node = node()</v>
        <v>PhaseArgs = term()</v>
        <v>Pid = pid()</v>
        <v>State = state()</v>
      </type>
      <desc>
        <p>This function is used to start an application with included
          applications, when there is a need for synchronization between
          processes in the different applications during startup.</p>
        <p>The start phases is defined by the application specification
          key <c>start_phases == [{Phase,PhaseArgs}]</c>. For included
          applications, the set of phases must be a subset of the set of
          phases defined for the including application.</p>
        <p>The function is called for each start phase (as defined for
          the primary application) for the primary application and all
          included applications, for which the start phase is defined.</p>
        <p>See <c>Module:start/2</c> for a description of
          <c>StartType</c>.</p>
      </desc>
    </func>
    <func>
      <name>Module:prep_stop(State) -> NewState</name>
      <fsummary>Prepare an application for termination</fsummary>
      <type>
        <v>State = NewState = term()</v>
      </type>
      <desc>
        <p>This function is called when an application is about to be
          stopped, before shutting down the processes of
          the application.</p>
        <p><c>State</c> is the state returned from
          <c>Module:start/2</c>, or <c>[]</c> if no state was returned.
          <c>NewState</c> is any term and will be passed to
          <c>Module:stop/1</c>.</p>
        <p>The function is optional. If it is not defined, the processes
          will be terminated and then <c>Module:stop(State)</c> is
          called.</p>
      </desc>
    </func>
    <func>
      <name>Module:stop(State)</name>
      <fsummary>Clean up after termination of an application</fsummary>
      <type>
        <v>State = term()</v>
      </type>
      <desc>
        <p>This function is called whenever an application has stopped.
          It is intended to be the opposite of <c>Module:start/2</c>
          and should do any necessary cleaning up. The return value is
          ignored.</p>
        <p><c>State</c> is the return value of
          <c>Module:prep_stop/1</c>, if such a function exists.
          Otherwise <c>State</c> is taken from the return value of
          <c>Module:start/2</c>.</p>
      </desc>
    </func>
    <func>
      <name>Module:config_change(Changed, New, Removed) -> ok</name>
      <fsummary>Update the configuration parameters for an application.</fsummary>
      <type>
        <v>Changed = [{Par,Val}]</v>
        <v>New = [{Par,Val}]</v>
        <v>Removed = [Par]</v>
        <v>&nbsp;Par = atom()</v>
        <v>&nbsp;Val = term()</v>
      </type>
      <desc>
        <p>This function is called by an application after a code
          replacement, if there are any changes to the configuration
          parameters.</p>
        <p><c>Changed</c> is a list of parameter-value tuples with all
          configuration parameters with changed values, <c>New</c> is
          a list of parameter-value tuples with all configuration
          parameters that have been added, and <c>Removed</c> is a list
          of all parameters that have been removed.</p>
      </desc>
    </func>
  </funcs>

  <section>
    <title>SEE ALSO</title>
    <p><seealso marker="doc/design_principles:des_princ">OTP Design Principles</seealso>,
      <seealso marker="kernel_app">kernel(6)</seealso>,
      <seealso marker="app">app(4)</seealso></p>
  </section>
</erlref>