--- layout: "docs" page_title: "Vagrant Triggers Configuration" sidebar_current: "triggers-configuration" description: |- Documentation of various configuration options for Vagrant Triggers --- # Configuration Vagrant Triggers has a few options to define trigger behavior. ## Execution Order The trigger config block takes two different operations that determine when a trigger should fire: * `before` * `after` These define _how_ the trigger behaves and determines when it should fire off during the Vagrant life cycle. A simple example of a _before_ operation could look like: ```ruby config.trigger.before :up do |t| t.info = "Bringing up your Vagrant guest machine!" end ``` Triggers can be used with [_actions_](#actions) or [_commands_](#commands) as well, but by default will be defined to run before or after a Vagrant guest. ## Trigger Options The trigger class takes various options. * `action` (symbol, array) - Expected to be a single symbol value, an array of symbols, or a _splat_ of symbols. The first argument that comes after either __before__ or __after__ when defining a new trigger. Can be any valid Vagrant command. It also accepts a special value `:all` which will make the trigger fire for every action. An action can be ignored with the `ignore` setting if desired. These are the valid action commands for triggers: - `destroy` - `halt` - `provision` - `reload` - `resume` - `suspend` - `up` * `ignore` (symbol, array) - Symbol or array of symbols corresponding to the action that a trigger should not fire on. * `info` (string) - A message that will be printed at the beginning of a trigger. * `name` (string) - The name of the trigger. If set, the name will be displayed when firing the trigger. * `on_error` (symbol) - Defines how the trigger should behave if it encounters an error. By default this will be `:halt`, but can be configured to ignore failures and continue on with `:continue`. * `only_on` (string, regex, array) - Limit the trigger to these guests. Values can be a string or regex that matches a guest name. * `ruby` (block) - A block of Ruby code to be executed on the host. The block accepts two arguments that can be used with your Ruby code: `env` and `machine`. These options correspond to the Vagrant environment used (note: these are not your shell's environment variables), and the Vagrant guest machine that the trigger is firing on. This option can only be a `Proc` type, which must be explicitly called out when using the hash syntax for a trigger. ```ruby ubuntu.trigger.after :up do |trigger| trigger.info = "More information" trigger.ruby do |env,machine| greetings = "hello there #{machine.id}!" puts greetings end end ``` * `run_remote` (hash) - A collection of settings to run a inline or remote script with on the guest. These settings correspond to the [shell provisioner](/docs/provisioning/shell.html). * `run` (hash) - A collection of settings to run a inline or remote script on the host. These settings correspond to the [shell provisioner](/docs/provisioning/shell.html). However, at the moment the only settings `run` takes advantage of are: + `args` + `inline` + `path` * `warn` (string) - A warning message that will be printed at the beginning of a trigger. * `exit_codes` (integer, array) - A set of acceptable exit codes to continue on. Defaults to `0` if option is absent. For now only valid with the `run` option. * `abort` (integer,boolean) - An option that will exit the running Vagrant process once the trigger fires. If set to `true`, Vagrant will use exit code 1. Otherwise, an integer can be provided and Vagrant will it as its exit code when aborting. ## Trigger Types Optionally, it is possible to define a trigger that executes around Vagrant subcommands and actions.