Adoptable Cookbooks List

Looking for a cookbook to adopt? You can now see a list of cookbooks available for adoption!
List of Adoptable Cookbooks

Supermarket Belongs to the Community

Supermarket belongs to the community. While Chef has the responsibility to keep it running and be stewards of its functionality, what it does and how it works is driven by the community. The chef/supermarket repository will continue to be where development of the Supermarket application takes place. Come be part of shaping the direction of Supermarket by opening issues and pull requests or by joining us on the Chef Mailing List.

Select Badges

Select Supported Platforms


storm-platform (9) Versions 1.5.0

Install/Configure a Storm cluster

cookbook 'storm-platform', '= 1.5.0'
cookbook 'storm-platform', '= 1.5.0', :supermarket
knife supermarket install storm-platform
knife supermarket download storm-platform
Quality 100%

Storm Cluster


Apache Storm is a free and open source distributed realtime computation system. Storm makes it easy to reliably process unbounded streams of data, doing for realtime processing what Hadoop did for batch processing. Storm is simple, can be used with any programming language, and is a lot of fun to use!


Cookbooks and gems

Declared in [metadata.rb](metadata.rb) and in [Gemfile](Gemfile).


A systemd managed distribution: - RHEL Family 7, tested on Centos

Note: it should work quite fine on Debian 8 (with some attributes tuning) but the official docker image does not allow Systemd to work easily, so it could not be tested.


Easy Setup

This cookbook creates the number of nodes specified by attribute node['storm-platform']['size'] (3 by default). The first node to be created will be setup with a Nimbus, a LogViewer and a UI. All other nodes will host a Supervisor and a LogViewer. To get this cookbook to setup the nth node to hold the Nimbus, Supervisor and LogViewer instead of the first one, change attribute node['storm-platform']['nimbus_id'].

By default, this cookbook installs openjdk from the official repositories (openjdk-headless 8 on centos 7) just before starting the service. You can change this behavior by setting node['storm-platform']['java'] to "", or choose your package by setting the package name in node['storm-platform']['java'][node[:platform]].

The recommended way to use this cookbook is through the creation of a role per Storm cluster. This enables the search by role feature, allowing a simple service discovery.

In fact, there are two ways to configure the search: 1. with a static configuration through a list of hostnames (attributes hosts that is ['storm-platform']['hosts']) 2. with a real search, performed on a role (attributes role and size like in ['storm-platform']['role']). The role should be in the run-list of all nodes of the cluster. The size is a safety and should be the number of nodes in the cluster.

If hosts is configured, role and size are ignored.

See [roles](test/integration/roles) for some examples and Cluster Search documentation for more information.

Zookeeper Cluster

To install properly a Storm cluster, you need a Zookeeper cluster. This is not in the scope of this cookbook but if you need one, you should consider using Zookeeper Platform.

The configuration of Zookeeper hosts use search and is done similarly as for Storm hosts, ie with a static list of hostnames or by using a search on a role. The attribute to configure is ['storm-platform']['zookeeper'].


This cookbook is fully tested through the installation of the full platform in docker hosts. This uses kitchen, docker and some monkey-patching.

If you run kitchen list, you will see 3 suites:

  • zookeeper-centos-7
  • storm-nimbus-centos-7
  • storm-supervisor-centos-7

Each corresponds to a different node in the cluster. They are connected through a bridge network named kitchen, which is created if necessary.

For more information, see [.kitchen.yml](.kitchen.yml) and [test](test) directory.

Local cluster

The cluster installed with kitchen converge is fully working and can thus be used as a local cluster for developments tests.

You can access it by using internal DNS of the docker network named kitchen or by declaring each node in your hosts file. You can get each IP by running:

docker inspect --format \
  '{{}}' container_name


Configuration is done by overriding default attributes. All configuration keys have a default defined in [attributes/default.rb](attributes/default.rb). Please read it to have a comprehensive view of what and how you can configure this cookbook behavior.



Install and fully configure a given Storm by including the recipes: install, create_user, config and systemd_service, in that order.


Install Storm from tar archive with ark cookbook.


Create a system user for Storm.


Configure Storm.


Create Systemd service files and set them up. Different Storm services will be set up following the kind of node (both may apply).

For a nimbus: - nimbus - logviewer - ui

For a supervisor: - supervisor - logviewer

Note: install java package (OpenJDK 8 on centos 7) by default, can be disabled by setting node['confluent-platform']['java'] to "". A platform specific configuration for the package to install is also possible.






Please read carefully []( before making a merge request.

License and Author

Copyright (c) 2015-2016 Sam4Mobile

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
See the License for the specific language governing permissions and
limitations under the License.

Dependent cookbooks

cluster-search >= 0.0.0
ark >= 0.0.0

Contingent cookbooks

There are no cookbooks that are contingent upon this one.




  • Use Storm 1.0.1 by default
  • Fix invalid check on cluster size
  • Use Continuous Integration with gitlab-ci


  • Fix Gemfile, need k-docker_cli instead of k-docker
  • Fix the too long line in kitchen_sandbox
  • Rename CHANGELOG -> and fix markdown


  • Fix: config could contain ruby/chef -specific yaml
  • Make Systemd unit path configurable, default is now "/etc/systemd/system"



  • Rename recipes containing '-' in their name (small breaking change):
    • create-user -> create_user
    • systemd-service -> systemd_service


  • Switch to docker_cli, use prepared docker image
    • Switch kitchen driver from docker to docker_cli
    • Use sbernard/centos-systemd-kitchen image instead of bare centos
    • Remove privileged mode :)
    • Remove some useless monkey patching
    • Use loops for suites generation
    • Remove dnsdock, use docker DNS (docker >= 1.10)
    • Use "kitchen" network, create it if needed


  • Fix rubocop offenses
  • Use specific names for resources to avoid cloning
  • Improve documentation, search config, etc.



  • Use storm 0.10.0 by default
  • Generate config in 2 steps: can use chef variables in erb style


  • Fix dirty yaml of nested structure for storm.yaml
  • Fix topology launch: create a link from storm.home/logs to log_dir
  • Fix auto-restart and also monitor log4j2 files
  • Fix: apply STORM-945 in default log4j configuration


  • Rationalize docker provision to limit images


  • Reorganize README:
    • Move changelog from README to CHANGELOG
    • Move contribution guide to
    • Reorder README, fix Gemfile missing
  • Add Apache 2 license file
  • Add missing chefignore


  • Strengthen tests reliability by sleeping after a Storm service is launch
  • Dependencies use supermarket, switch to zookeeper-platform


  • Cleaning, use java-1.8.0-openjdk-headless as default java package


  • Initial version with Centos 7 support

Foodcritic Metric

1.5.0 passed this metric