mozilla-services/elmo-automation

Name: elmo-automation

Owner: Mozilla Services

Description: docker-compose setup for a10n, master-ball, slave-ball, etc.

Created: 2017-02-02 16:35:52.0

Updated: 2018-05-04 12:46:51.0

Pushed: 2018-05-04 12:46:50.0

Homepage: null

Size: 28748

Language: Python

GitHub Committers

UserMost Recent Commit# Commits

Other Committers

UserEmailMost Recent Commit# Commits

README

Elmo Automation

This repository contains the automation behind https://l10n.mozilla.org, aka Elmo. It's built on top of docker compose, and is deployed on SCL3 and AWS.

Please use bugzilla to file issues/bugs.

Find more information on Elmo itself in its repo.

Volumes

There's a shared data volume between some of the containers below. It's used to cache bare repository data. That can be used in containers to access that, or to hg share repositories cheaply to local disk.

Containers

For testing only, there's an hg container. It mocks https://hg.mozilla.org/, in a rather naive way, though. As we test the automation, this is where the data flow starts.

hg-poller

This container scrapes hg for new pushes, by continuously loading the json-pushes endpoint for each known repository. It feeds newly found pushes to rabbitmq.

a10n-hg-worker

This container observes the queue that hg-poller writes to, and syncronizes the new push data with the elmo database. To do so, it also updates the shared clones in order to do so.

bb

This container runs buildbot 0.7.x. It runs the master, which is responsible for polling the elmo database for new changes, schedule builds, and gather status and logs. The status is mirrored in the elmo database.

It also runs the slave, which is doing the actual work. The build steps contain

The slave stores a summary of the compare-locales result in the elmo database, and uploads the details as a JSON blob to es.

es and rabbitmq

These two are stock instances of ElasticSearch and RabbitMQ.


This work is supported by the National Institutes of Health's National Center for Advancing Translational Sciences, Grant Number U24TR002306. This work is solely the responsibility of the creators and does not necessarily represent the official views of the National Institutes of Health.