Rand Stats




Sparrow Based Configuration Management Tool.

Build status

Build Status


zef install Sparrowdo


sparrowdo --host=remote-host [options]

For example:

# ssh host, bootstrap first, use of local repository
sparrowdo --host=  --repo=file:///home/gimy/repo --verbose  --bootstrap

# docker host, don't add sudo
sparrowdo --docker=SparrowBird --no_sudo

# localhost, with sudo
sparrowdo --localhost

# read hosts from file
sparrowdo --host=hosts.raku


Sparrow6 DSL is high level functions to manage configurations.

Just create a file named sparrowdo in current working directory and add few functions:

$ cat sparrowfile

user 'zookeeper';

group 'birds';

directory '/var/data/avifauna/greetings/', %( owner => 'zookeeper' );

file-create '/var/data/avifauna/greetings/sparrow.txt', %( 
    owner => 'zookeeper',
    group => 'birds', 
    mode  => '0644', 
    content => 'I am little but I am smart'

service-start 'nginx';

package-install ('nano', 'ncdu', 'mc' );

The full list of function is available at Sparrow6 DSL documentation.

If DSL is not enough you can call any Sparrow6 plugin using task-run function:

task-run "show me your secrets", "azure-kv-show", %(
  kv      => "Stash" # key vault name
  secret  => ( "password", "login" ) # secret names

DSL vs task-run function

DSL is high level adapter with addition of some "syntax sugar" to make a code concise and readable.

DSL functions are Perl6 functions, you take advantage of input parameters validation.

However DSL is limited. Not every Sparrow6 plugin has related DSL function.

Once we've found some Sparrow6 plugin common enough we add a proper DSL wrapper for it.

In case you need more DSL wrappers let us know!

Sparrowdo workflow

  ====================                                   ====================
  |                  |  --- [ Bootstrap ]                |                  | LocalHost
  |    MasterHost    |  --- [ Sparrowdo Scenario ] --->  |    TargetHost    | Docker 
  |                  |               *                   |                  | Ssh
  ====================               |                   ====================
   {Sparrow6 Module}                 |                    {Sparrow6 Module}
                                      \    parameters
                                       \-----------------> [ tasks ]
                                                            ^  ^  ^
                                                           /  /  /
                             ------------                 /  /  /
                            { Repository } ---------------------
                             ------------        [ plugins ]
                            /      |     \
                           /       |      \
                          /        |       \
                         /         |        \
                      [CPAN]   [RubyGems] [PyPi]

Master host

Master host is where Sparrow6 tasks "pushed" to target hosts.

Sparrowdo should be installed on master host:

$ zef install Sparrowdo

Sparrowdo scenario

Sparrowdo scenario is collection of Sparrow6 tasks executed on target host.

Target hosts

Target hosts are configured by Sparrow6 which should be installed there as well:

$ zef install Sparrow6


Bootstrap is process of automation of Sparrow6 installation on target host:

$ sparrowdo --host= --bootstrap

Bootstrap runs under sudo and requires sudo privileges on target host.


Execution of Sparrow6 tasks on target host is made under sudo by default, to disable this use no_sudo flag:

$ sparrowdo --docker=brave_cat --no_sudo # We don't need sudo on docker

Various types of target hosts

Once sparrowfile is created it could be run on target host.

Ssh host:

$ sparrowdo --host=

Docker container:

$ sparrowdo --docker=old_dog


$ sparrowdo --localhost


Sparrowdo use Sparrow6 repository to download tasks that are executed on target hosts.

During scenario execution Sparrowdo pulls plugins from repository and runs it with parameters.

Plugin that gets run with parameters is a Sparrow6 task.

By default local repository located at ~/repo is used:

$ sparrowdo --host= #  using repo ~/repo

Use --repo flag to set repository URI. For example:

$ sparrowdo --repo=file:///tmp/repo --localhost # use /tmp/repo repository for local mode

$ docker run --name --name=fox -d sh -v /tmp/repo:/var/data

$ sparrowdo --repo=/var/data --docker=fox --repo=file://var/data # mount /tmp/repo to /var/data within docker instance

$ sparrowdo --repo=http://sparrow.repo # use http repository

Ssh mode supports synchronizing local file repositories with master and target host using scp. It extremely useful if for some reason you can't use http based repositories:

# scp local repository /tmp/repo to target host
# and use it within target host

$ sparrowdo --sync=/tmp/repo --host=

Read Sparrow6 documentation to know more about Sparrow6 repositories.

Sparrowdo files anatomy

When Sparrowdo scenario executed there are might be related that get copied to a target host, for convenience, so that one can reliably refer to those files inside Sparrowdo scenario.

Thus Sparrowdo project might consists of various files and folders:

├── conf
│   └── alternative.pl6
├── config.pl6
├── data
│   └── list.dat
├── files
│   └── baz.txt
├── sparrowfile
└── templates
    └── animals.tmpl

The minimal setup is just sparrowfile placed in the current working directory.

That file contains Sparrowdo scenario to be executed.

One may optionally define other data that make up the whole set up:

Directory to hold configuration files

Default configuration file, if placed in current working directory will define Sparrowdo configuration, see also scenarios configuration section, alternatively one may place configuration file to conf directory and choose to run with that file:

$ sparrowdo --conf=conf/alternative.pl6

Directory containing files, so one could write such a scenario:

file "/foo/bar/baz.txt", %( source => "files/baz.txt" )

The same as files directory, but for templates, technically templates and files as equivalent, but named separately for logical reasons. Obviously one can populate templates using that directory:

template-create '/var/data/animals.txt', %(
  source => ( slurp 'templates/animals.tmpl' ),
  owner => 'zookeeper',
  group => 'animals' ,
  mode => '644',
  variables => %(
    name => 'red fox',
    language => 'English'

The same as files and templates folder but to keep "data" files, just another folder to separate logically different pieces of data:

file "/data/list.txt", %( source => "data/list.dat" )

Scenario configuration

If config.pl6 file exists at the current working directory it will be loaded via Perl6 EVALFILE at the beginning of scenario.

For example:

$ cat config.pl6

  user         => 'foo',
  install-base => '/opt/foo/bar'

Later on in the scenario you may access config data via config function, that returns configuration as Perl6 Hash object:

$ cat sparrowfile

my $user         = config<user>;
my $install-base = config<install-base>;

To refer configuration files at arbitrary location use --conf parameter:

$ sparrowdo --conf=conf/alternative.conf

See also Sparrowdo files anatomy section.

Including other Sparrowdo scenarios

Because Sparrowdo scenarios are Perl6 code files, you're free to include other scenarios into "main" one using via Perl6 EVALFILE mechanism:

$ cat sparrowfile

EVALFILE "another-sparrowfile";

Including another scenario does not include related files (data, files, templates) into main one, so you can only reuse Sparrowdo scenario itself, but not it's related files.

Bootstrap and manual install

To ensure that target hosts has all required Sparrow6 dependencies one may choose bootstrap mechanism that installs those dependencies on target host. This operation is carried out under sudo.

Bootstrap is disabled by default and should be enabled by --bootstrap option:

$ sparrowdo --host= --bootstrap

Be aware that all dependencies will be installed system wide rather then user account, because sudo is used.

if you don't want bootstrap target host, you can always choose to install Sparrow6 zef module on target host, which is enough to start using Sparrowdo to manage target host:

$ zef install Sparrow6

Run Sparrowdo on docker

Sparrowdo supports docker as target host, just refer to running docker container by name to run Sparrowdo scenario on it:

$ sparrowdo --docker=running-horse

Run Sparrowdo on ssh hosts

This is probably one the most common use case for Sparrowdo - configuring servers by ssh:

$ sparrowdo --host=

Run Sparrowdo in local mode

In case you need to run sparrowdo on localhost use --localhost flag:

$ sparrowdo --localhost

Sparrowdo cli

sparrowdo is Sparrowdo cli to run Sparrowdo scenarios, it accepts following arguments:

One of two things:

Ssh host to run Sparrowdo scenario on. Default value is For example:


Path to raku file with dynamic hosts configuration. For example:


This will effectively runs Sparrowdo in asynchronous mode through a Sparky backend. See Dynamic hosts wiki.

Docker container name to run Sparrowdo scenario on, none required.

If set, runs Sparrowdo in localhost, in that case Sparrow6 tasks get run directly without launching ssh

If set defines ssh user when run Sparrowdo scenario over ssh, none required.

If set defines ssh port to use when run Sparrowdo scenario over ssh. Default value is 22.

If set, run Sparrowdo scenario on localhost, don't confuse with --host= which runs scenario on the machine but using ssh.

Defines Sparrow6 repository URI, for example:

--repo= # Remote http repository

--repo=file:///var/data/repo # Local file system repository

If --repo is not set, Sparrowdo uses default local repository located at ~/repo path.

Sparrowdo uses repository to pull plugins from, see also Sparrowdo workflow section.

See also Sparrow6 Repositories documentation.

If set synchronize local file system repository with target hosts using scp, it's useful if for some reasons remote http repositories are not available or disable, so one can automatically copy repository from master host to target. --sync parameter value should file path pointing repository directory at master host:


If set, don't request Sparrow6 repository update. Useful when in limited internet and using http based remote repositories.

Runs in dry run mode. Only applies when run in asynchronous mode through a Sparky backend.

See Dynamic hosts wiki.

Set host(s) tags, see Dynamic hosts wiki.

If set print additional information during Sparrowdo scenario execution.

If set, runs Sparrow6 tasks in debug mode, giving even more information, during scenario execution.

If set, define path to be appending to local cache directory on target host, used predominantly when running concurrent Sparrowdo scenario on one target host, for example in Sparky CI server. None required, no default value.

Advanced topics

Run Sparrowdo scenarios in asynchronous mode, using Sparky backend


Alexey Melezhik

See also


To God as the One Who inspires me to do my job!