Skip to main content

Command-line client for Gilliam

Project description

# The Gilliam Command-Line Interface

Gilliam is a software platform for building and running distributed

This is the command-line tool that is used to control services running
on the Gilliam platform.

## Installation

Install using pip (preferably in a virtualenv) straight from the

$ pip install git+

That's it. The command is called `gilliam`.

# Quick Intro

First a few words about Gilliam application model: A *service* is a
piece of code that runs on the platform. It can be some business
logic, a cache or a database. Services are grouped together into a
*formation*. Code and configuration in the form of environment
variables are merged together into a *release*. *Instances* of a
service can be created from a release. Everything live in a *stage*,
which is another word for a gilliam installation.

In this quick into we'll use a small python example to showcase some
of the commands of the client. First, we need to clone the example

$ git clone git+
$ cd python-example

Lets take a look at the `gilliam.yml` file:

script: python
ports: [80]

A `gilliam.yml` file specifies a set of services that the formation
need to expose functionality. In file above contains just a process
called `www`. The `script` tells us that when the service is started
the command `python` should be executed. `ports` specifies
ports on which the service can receive incoming requests.

Gilliam comes with a whole set of different service types. The type
for a service can always be specified with a `type` field. But you
can also let it to Gilliam to figure out the type; the type will then
be deduced from the service name and its definition.

Back to the example.

To create a formation for the example issue the following command:

$ gilliam --stage myapp-prod create python-example

We specify `--stage myapp-prod` to tell the client to create the
formation in the `myapp-prod` stage (the production environment for my

*Note*: This quick into does not take into consideration how the stage
is created.

Next step is to build and deploy:

$ gilliam deploy
start building service 'www' ...
| Python app detected
| -----> No runtime.txt provided; assuming python-2.7.4.
| Successfully installed Flask Werkzeug Jinja2 markupsafe
| Cleaning up...
| Cleaning up caches
release 1

When done, a new release has been created. You can see it if you do a
`gilliam-cli releases`:

$ gilliam releases
name author message
--------- --------------- ----------------------------------------
1 jrydberg None

Gilliam allows heterogeneous set ups where you have instances from
multiple releases running at the same time. This allows you to build
sophisticated build pipelines with canary tests.

But right now, nothing is running, since we have not scaled the
release. So lets do that:

$ gilliam scale -r1 www=1

This will create an instance of the `www` service and dispatch it to
one of the executors. Run `gilliam-cli ps` to inspect instances:

$ gilliam-cli ps
name release state
----------------------------------- ------- ---------
www.BHCrBMebfG4oZUgix95chH 1 running

To be able to access the service from the outside, we need to set up
a route:

$ gilliam-cli route /example/ www.example.service
route 5h7Zf9P3oDGiUpF3opWDzP created

So what does that mean? It means that `/example/` on the router,
regardless of domain, will route requests to the example service.

Routes can be listed by issuing `routes` command:

$ gilliam-cli route
name domain path target
---------------------- -------------------- -------------------- -------------------------
5h7Zf9P3oDGiUpF3opWDzP /example/ http://www.example.service

# Basic Commands

## Creating a Formation

The first command you will issue. It needs to be issued from your
"project dir", which is the directory where your `gilliam.yml` file
live. Normally this is the root directory of your project.

$ gilliam --stage prod create name-of-formation

The `--stage` option tells gilliam in what installation the formation
should be created. Here's it's the production environment.

## Deploying

To build your project and release it just issue the `deploy` command:

$ gilliam deploy

## Scaling a Release

Scaling a release means to increase or decrease the number of
instances of that release.

(FIXME: default to last release, or to <all releases> somehow?)

## Routing

Gilliam has a front-end HTTP router that takes request and forwards
them to a service for process.

The command `route` takes two arguments, the *route* and the *target*.

$ gilliam route /example/ www.example.service

The route can contain a domain name that needs to be matched:

$ gilliam route www.example.service

The route argument accepts variable matching using the `{var}` syntax,
like this:

$ gilliam route /user/{user} www.user.service/{user}

If you want to specify a specific format for the variable, do so after
a colon, like this: `{name:REGEX}`. For example `{rest:.*?}` will
match the rest of the line, which can be useful if you want to send
everything to a specific service.

$ gilliam route /user/{rest:.*?} www.user.service/{rest}

Note that variable matching is also possible on the domain:

$ gilliam route {acct}{rest:.*?} api.user.service/{acct}/{rest}

Or in the formation name:

$ gilliam route {acct}{rest:.*?} api.user-{acct}.service/{rest}

By not specifying any arguments all existing routes will be listed:

$ gilliam routes
name domain path target
---------------------- -------------------- -------------------- -------------------------
5h7Zf9P3oDGiUpF3opWDzP /example/ http://www.example.service

Deleting a formation is done using the `-d` option:

$ gilliam route -d <route-name>

## Building a Release


Project details

Release history Release notifications

This version
History Node


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
gilliam-cli-0.1.tar.gz (26.1 kB) Copy SHA256 hash SHA256 Source None

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page