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

Select Status

RSS

storm-platform (9) Versions 2.1.0

Install/Configure a Storm cluster

Policyfile
Berkshelf
Knife
cookbook 'storm-platform', '~> 2.1.0', :supermarket
cookbook 'storm-platform', '~> 2.1.0'
knife supermarket install storm-platform
knife supermarket download storm-platform
README
Dependencies
Changelog
Quality 50%

Storm Cluster

Description

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.

This cookbooks installs and configures Storm (> 1.0.0).

Requirements

Cookbooks and gems

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

Platforms

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.

Usage

First Setup

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']) (empty by default so deactivated)
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 and search is
deactivated.

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

By default, the first node in hosts array or in search result (sorted on fqdn)
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 be
the Nimbus instead of the first one, change attribute
node['storm-platform']['nimbus_id']. You can also activate Nimbus High
Availability. You will find more information in the following sections.

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 are located in
node['storm-platform']['zookeeper'].

Java

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 "" (empty
string), or choose your package by setting the package name in
node['storm-platform']['java'][node[:platform]].

Nimbus High Availability

A new feature of Storm 1.0.0 is Nimbus High Availability. This allows a backup
nimbus to be elected as master nimbus when the later crashes. To do so, you
have to declare all possible nimbus. This can be done by setting
node['storm-platform']['nimbus_id'] (default: 1) to the first nimbus and
node['storm-platform']['n_of_nimbus'] (default: 1) to the number of nimbus.

Test

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 4 suites:

  • zookeeper-centos-7
  • storm-01-centos-7
  • storm-02-centos-7
  • storm-03-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 \
  '{{.NetworkSettings.Networks.kitchen.IPAddress}}' container_name

Attributes

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.

Recipes

default

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

install

Install Storm from tar archive with ark cookbook.

create_user

Create a system user for Storm.

config

Configure Storm.

systemd_service

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

If multiple nimbus exist (to activate nimbus High Availability) then we
consider each nimbus to be both nimbus and supervisor. In this case, a
nimbus will run all the services.

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

Resources/Providers

None.

Changelog

Available in [CHANGELOG](CHANGELOG).

Contributing

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

License and Author

Copyright (c) 2015-2017 Sam4Mobile, 2018 Make.org

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

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

Dependent cookbooks

ark >= 0.0.0
cluster-search >= 0.0.0

Contingent cookbooks

There are no cookbooks that are contingent upon this one.

Changelog

2.1.0

Main:

  • chore: handover maintenance to make.org
  • feat: use 1.2.2 as default Storm version
  • fix: set archive url to always get old version

Tests:

  • test: include .gitlab-ci.yml from test-cookbook
  • test: replace deprecated require_chef_omnibus
  • test: reduce memory usage, increase timeout (huge)

Misc:

  • chore: set new contributing guide with karma style
  • chore: add foodcritic & rubocop in Gemfile
  • style(rubocop): fix misc offenses like delimiters
  • style: use alias cookbook_name everywhere
  • refactor: fix deprecation warning on chef_gem[xml-simple]
  • fix: fix shell for storm user

2.0.0

Main:

  • Require Storm > 1.0.0 (because of HA, nimbus.host is no longer valid)
  • Configure nimbus HA by allowing multiple nimbus (with n_of_nimbus key)
  • Fix default config for Storm > 1.0.0
  • Fetch log4j2 default config directly from github
  • Do not wait anymore after (re)starting a service

Test:

  • Strengthen tests: test all nodes and check deactivated services
  • Smart sleeps to wait services to be up

Misc:

  • Improve documentation (better explanation of search, add HA, etc.)

1.5.0

Main:

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

Misc:

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

1.4.0

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

1.3.0

Main:

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

Test:

  • 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

Misc:

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

1.2.0

Main:

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

Fixes:

  • 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

Test:

  • Rationalize docker provision to limit images

Packaging:

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

1.1.0

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

1.0.1

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

1.0.0

  • Initial version with Centos 7 support

Collaborator Number Metric
            

2.1.0 passed this metric

Contributing File Metric
            

2.1.0 failed this metric

Failure: To pass this metric, your cookbook metadata must include a source url, the source url must be in the form of https://github.com/user/repo, and your repo must contain a CONTRIBUTING.md file

Foodcritic Metric
            

2.1.0 passed this metric

No Binaries Metric
            

2.1.0 passed this metric

Testing File Metric
            

2.1.0 failed this metric

Failure: To pass this metric, your cookbook metadata must include a source url, the source url must be in the form of https://github.com/user/repo, and your repo must contain a TESTING.md file

Version Tag Metric
            

2.1.0 failed this metric

Failure: To pass this metric, your cookbook metadata must include a source url, the source url must be in the form of https://github.com/user/repo, and your repo must include a tag that matches this cookbook version number