bbc/hive-scheduler

Name: hive-scheduler

Owner: BBC

Description: null

Created: 2015-07-10 14:31:16.0

Updated: 2017-01-12 08:48:57.0

Pushed: 2017-10-20 11:40:12.0

Homepage: null

Size: 1603

Language: Ruby

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

Hive Scheduler

The scheduler component behind the hive, and the main web view. For a full list of release notes, see the change log

Installing

Pre-requisites

• git - Some of the gems are only currently available via Github
• MySQL client libraries - sudo apt-get install libmysqlclient-dev on Ubuntu
• A javascript runtime - For example, NodeJS - sudo apt-get install nodejs on Ubuntu

Configuration

Hive-Scheduler is configured using the Chamber gem. You can add configuration directly into the config/settings.yml file, or by setting the following environment variables (in the following lines to ~/.bashrc (or equivalent):

rt RAILS_ENV=production
rt DATABASE_ADAPTER=mysql2
rt DATABASE_ENCODING=utf8
rt DATABASE_HOST=your_database_host
rt DATABASE_PORT=3306
rt DATABASE_POOL=5
rt DATABASE_USERNAME=database_username
rt DATABASE_PASSWORD=database_password
rt DATABASE_DATABASE=database_name
rt ATTACHMENT_STORAGE=filesystem
rt ATTACHMENT_STORAGE_PATH_BASE=public
rt HIVE_QUEUES=true

Set up database

n/rake db:migrate
n/rake db:seed
n/rake assets:precompile

Example data

Some example execution scripts and projects can be created by editing the db/seeds.rb file before executing the rake tasks above.

Start the server

n/rails s

Examples

Hello world

Go into the 'Scripts' section and create a new script. Select 'Shell Script' as the target platform. Set the name to 'Hello ' and in the template box enter:

is will use the execution variable 'word'
 Hello $HIVE_WORD

Add a new execution variable called word and set the field type to 'String'. Save the script.

Go into the 'Projects' section and create a new project. Set the name to 'Hello world' and select the 'Hello ' script. Select the 'Manual' population mechanism and enter 'world' in the Word field and 'bash' in the Queues field. Leave all other fields as the defaults. Save the project.

Go into the 'Batches' section and create a new batch. Set the name to 'First test batch'. Select 'Hello world' as the project and set the version to '1.0'. Leave all other fields as the defaults. Save the batch.

In the 'Batches' section you will now see the batch you have just created and by clicking on the name you can see a single job belonging to this batch for the queue 'bash'. If you have a hive set up to run shell tests on a queue named 'bash' then this test will be executed and by clicking on the job number you can view the output.

Android test

The Android Calabash script is set up by default. In the 'Projects' section create a new project and select 'Android Calabash' as the execution type. Set the name to 'X Platform Example' and enter the following in the repository field:

github.com:calabash/x-platform-example.git

Select 'Manual' as the population mechanism and enter 'android-5.1.1' in the queue field. Click on the 'Add queue' link and enter 'android-4.4.4' in the new queue field.

Note; the hive runner is configured to run tests listed in queues specified by name. The values entered in these fields should be changed as appropriate for the configuration of your hives.

Save the project.

In the 'Batches' section create a new batch and select the project 'X Platform Example'. Set the name to 'Testing X platform example' and set the version. Upload the apk, which can be found at

• https://github.com/calabash/x-platform-example/tree/master/prebuilt

Leave all other fields unchanged (the list of queues may be amended as required) and save the batch. As with the 'Hello world' example, the jobs will run if hives are set up with devices for the given queues and output can be viewed.

License

Hive Scheduler is available to everyone under the terms of the MIT open source licence. Take a look at the LICENSE file in the code.

Copyright (c) 2015 BBC