Every command under commands in vagga.yaml is mapped with a tag that denotes the command type. The are two command types !Command and !Supervise illustrated by the following example:

containers: {ubuntu: ... }
  bash: !Command
    description: Run bash shell inside the container
    container: ubuntu
    run: /bin/bash
  download: !Supervise
    description: Download two files simultaneously
      amd64: !Command
        container: ubuntu
        run: wget
      i386: !Command
        container: ubuntu
        run: wget

Since 0.7.1 there is a third experimental kind which is !CapsuleCommand.

Common Parameters

These parameters work for both kinds of commands:


Description that is printed in when vagga is run without arguments


The message that is printed before running process(es). Useful for documenting command behavior.


The seconds to sleep before printing banner. For example if commands run a web service, banner may provide a URL for accessing the service. The delay is used so that banner is printed after service startup messages not before. Note that currently vagga sleeps this amount of seconds even if service is failed immediately.


The message printed after command is run. It’s printed only if command returned zero exit status. Useful to print further instructions, e.g. to display names of build artifacts produced by command.


The list of commands to run before the command, each time it is started.


    container: build
    run: "make prog"
    container: build
    prerequisites: [make]
    run: "./prog"

The sequence of running of command with prerequesites is following:

  1. Container is built if needed for each prerequisite
  2. Container is built if needed for main command
  3. Each prerequisite is run in sequence
  4. Command is started

If any step fails, neither next step nor the command is run.

The prerequisites are recursive. If any of the prerequisite has prerequisites itself, they will be called. But each named command will be run only once. We use topology sort to ensure prerequisite commands are started before dependent commands. For cyclic dependencies, we ensure that command specified in the command line is run later, otherwise order of cyclic dependencies is unspecified.

The supervise command’s --only and --except influences neither running prerequisites itself nor commands inside the prerequisite if the latter happens to be supervise command. But there is a global flag --no-prerequisites.

The prerequisites is not (yet) supported in the any of children of a !Supervise command, but you can write prerequisites for the whole command group.


Check the sysctl fs.inotify.max_user_watches and print a warning or set it automatically if auto-apply-sysctl is enabled. More info about max_user_watches


This is a docopt definition for the options that this command accepts. Example:

  test: !Supervise
    options: |
      Usage: vagga test [--redis-port=<n>] [options] [<tests>...]

        -R, --redis-port <n>  Port to run redis on [default: 6379]
        <tests> ...           Name of the tests to run. By default all
                              tests are run
      redis: !Command
        container: redis
        run: |
          redis-server --daemonize no --port "$VAGGAOPT_REDIS_PORT"
      first-line: !Command
        container: busybox
        run: |
          py.test --redis-port "$VAGGAOPT_REDIS_PORT" $VAGGAOPT_TESTS

As you might have noticed, options are passed in environment variables prefixed with VAGGAOPT_ and VAGGACLI_ (see below). Your scripts are free to use them however makes sense for your application.


  • You should include [options] at least in one of the usage examples, to have -h, --help working as well as other built-in options (--only, --except in supervise commands)
  • This setting overrides accepts-arguments

Every argument is translated into two variables:

  • VAGGAOPT_ARG – has the raw value of the argument, for boolean flags it contains either true or nothing, for repeatable flags it contains a number of occurences
  • VAGGACLI_ARG – has a canonical representation of an argument, this includes option name and all needed escaping to represent multiple command line arguments

The ARG is usually a long name of the option if exists, or short name otherwise. For positional arguments it’s argument name. It’s always uppercased and has - replaced with _

There are few shortcommings of both kinds:

  1. VAGGAOPT_ can’t represent list of arguments that can contain spaces. So it can’t be used for list of file names in the general case.
  2. VAGGACLI_ contains escaped versions of arguments so requires using eval to make proper argument list from it

Some shell patterns using VAGGAOPT_:

  1. To propagate a flag, use either one:

    somecmd ${VAGGAOPT_FLAG:+--flag}
    somecmd $VAGGACLI_FLAG
  2. To optionally pass a value to a command, use either one (note the implications of eval in the second command):

    somecmd ${VAGGAOPT_VALUE:+--value} $VAGGAOPT_VALUE
    eval somecmd $VAGGACLI_VALUE

    To overcome limitations of eval, for example if you need to expand $(hostname) in the command, you can use the following snippet:

    eval printf "'%s\0'" $VAGGACLI_VALUE | xargs -0 somecmd -H$(hostname)
  3. To pass a list of commands each prefixed with a --test=, use either one:

    # any shell (but ugly)
    eval printf "'%s\0'" $VAGGACLI_TESTS | sed -z 's/^/--test=/' | xargs -0 somecmd
    # bash only
    eval "tests=($VAGGACLI_TESTS)"
    somecmd "${tests[@]/#/--test=}"

    (Note for some sed implementations you need to omit -z flag)

    This works if you have argument like vagga test <tests>.... However, if your vagga command-line is vagga test --test=<name>... use the following instead:

    eval somecmd $VAGGACLI_TEST

This functionality is experimental. Some details can change in future.

If parameter is set to some name, then vagga _update_symlinks will add a symlink to a vagga binary with this name. This should be used for commands that you want to run from your shell without vagga␣ prefix. See _update_symlinks command for more info.

New in version 0.7.1.

Parameters of !Command


The container to run command in.


The list of tags for this command. Tags are used for processes filtering (with --only and --exclude) when running any !Supervise command.

Simple example:

  run: !Supervise
      postgres: !Command
        tags: [service]
        run: ...
      redis: !Command
        tags: [service]
        run: ...
      app: !Command
        tags: [app]
        run: ...
$ vagga run --only service  # will start only postgres and redis processes

The command to run. It can be:

  • either a string encompassing a shell command line (which is feeded to /bin/sh -c)
  • or a list containing first the full path to the executable to run and then possibly static arguments.

The working directory to run in. Path relative to project root. By default command is run in the same directory where vagga started (sans the it’s mounted as /work so the output of pwd would seem to be different)


Denotes whether command accepts additional arguments. Defaults to:

  • false for a shell command line (if run is a string);
  • true if command is an executable (if run is a list).

NB: If command is a shell command line - even if it’s composed of only one call to an executable -, arguments are given to its executing context, not appended to it.


This setting is ignored when options is set.


The mapping of environment to pass to command. This overrides environment specified in container on value by value basis.


The mapping of mount points to the definition of volume. Allows to mount some additional filesystems inside the container. See Volumes for more info.

The volumes defined here override volumes specified in the container definition (each volume name is considered separately).


You must create a folder for each volume. See Container Building Guide for documentation.


This denotes what is run as pid 1 in container. It may be wait, wait-all-children or exec. The default wait is okay for most regular processes. See What’s Special With Pid 1? for more info.


The parameter specifies how container’s base file system is used. By default container is immutable (corresponds to the read-only value of the parameter), which means you can only write to the /tmp or to the /work (which is your project directory).

Another option is transient-hard-link-copy, which means that whenever command is run, create a copy of the container, consisting of hard-links to the original files, and remove the container after running command. Should be used with care as hard-linking doesn’t prevent original files to be modified. Still very useful to try package installation in the system. Use vagga _build --force container_name to fix base container if that was modified.


The user id to run command as. If the external-user-id is omitted this has same effect like using sudo -u inside container (except it’s user id instead of user name)


(experimental) This option allows to map the user-id as seen by command itself to some other user id inside container namespace (the namespace which is used to build container). To make things a little less confusing, the following two configuration lines:

user-id: 1
external-user-id: 0

Will make your command run as user id 1 visible inside the container (which is “daemon” or “bin” depending on distribution). But outside the container it will be visible as your user (i.e. user running vagga). Which effectively means you can create/modify files in project directory without permission errors, but tar and other commands which have different behaviour when running as root would think they are not root (but has user id 1)


The group id to run command as. Default is 0.


The list of group ids of the supplementary groups. By default it’s empty list.


Binds a TCP to the specified address and passes it to the application as a file descriptor #3.


  container: nginx
  run: nginx
  pass-tcp-socket: 8080
    NGINX: "3;" # inform nginx not to listen on its own

You may specify what to listen to with the following formats:

  • 8080 – just a port number – listens on
  • *:8080 – wildcard pattern for host – listens on every host
  • – same as *:8080
  • – listen on specified IPv4 host
  • [2001:db8::1]:8080 – listen on specified IPv6 host
  • localhost:8080 – resolve a name and listen that host (note: name must resolve to a single address)

This is better then listening by the application itself in the following cases:

  1. If you want to test systemd socket activation
  2. If you prepare your application to a powerful supervisor like lithos (lithos can run multiple processes on the same port using the feature)
  3. To declare (document) that your application listens specified port (otherwise it may be hidden somewhere deeply in config)
  4. To listen port in the host network namespace when applying network isolation (as an alternate to public-ports)

Parameters of !Supervise


The set of processes to supervise and mode. See Supervision for more info


A mapping of name to child definition of children to run. All children are started simultaneously.


(default 2 seconds) If some process exits (in stop-on-failure mode), vagga will send TERM signal to all the other processes. If they don’t finish in the specified number of seconds, vagga will kill them with KILL signal (so they finish without being able to intercept signal unconditionally). If you don’t like this behavior set the parameter to some large value.


Run all processes within isolated network namespace. Isolated network will have only a loopback device, so processes won’t have access to the internet. For example, it is possible to run several test suites each start service that binds the same port. Also you can run arbitrary command inside isolated network using --isolate-network commandline option.

Parameters of !CapsuleCommand

This functionality is experimental. Some details can change in future.

It’s generally not recommended to use, unless you know what are you doing.


If you need to build container non-standard uid/gid mapping, you might want to configure these to match container’s settings.

Note: all containers built in this CapsuleCommand must match uids/gids with these. Defaults are usually okay.


The working directory to run in. Path relative to project root. By default command is run in the same directory where vagga started (sans the it’s mounted as /work so the output of pwd would seem to be different)


The mapping of environment to pass to command. This overrides environment specified in container on value by value basis.

This is also a good way to pass some variables to a script that bootstraps a capsule.


The command to run. Rules are the same as in normal command, but usually this option either specifies a local script to run (committed in repository) or a single pattern like vagga _capsule script https://some/ See Capsule Commands for more info.