Auto-tools/Projects/WebRTC Cloud Automation

This is our project page for the WebRTC/Services Automation system that we're building in the cloud. Our first target is to build something in AWS and then to build out into hardware since we can't test everything we'd like to test in AWS. However, we should be able to use the same configuration, tests, and set-up for both types of testing.

Design

Phase 1

• Prototype system in AWS, using VPC, NAT, TURN and STUN servers
• Create test runner (ted)
• Create execution environment on client (dminor)
• Define test format, get test running (adam)
  • Support both C++ as well as web page tests?
• Milestone 1: test runner + tests coordinating, running tests on separate machines

Phase 1 AWS network diagram:

Phase 2

• Automatically being able to specify a network configuration in the test and create/remove components before test
  • Perhaps borrow from Releng's existing items:
    • RelEng is using Boto to communicate with AWS instances (https://github.com/boto/boto)
    • Start script: http://hg.mozilla.org/build/cloud-tools/file/5e627ea3e624/aws/aws_watch_pending.py
    • Stop script: http://hg.mozilla.org/build/cloud-tools/file/5e627ea3e624/aws/aws_stop_idle.py
    • Rail says Boto is working well for them, other than the fact it is a synchronous interface

M1:Q3

• EC2 machines, automated fashion by the end of Q3 (not on check-in)
  • Roughly a check-in every 20 minutes (this will take more of a cluster of machines
  • Any test that takes less than 20 minutes to run should be kosher; we can shoot for that but we don't want to shoot too high and get burned
  • Initial work to get Jenkins going is here: http://54.215.197.13:8080/

M2:Q4

• RelEng supported environment
• Different Network hardware and testing on the above

Meetings

Held every Tuesday at 12:30pm Pacific Time, 3:30pm Eastern.

* Ted's vidyo Room
* Dial in: 1-800-707-2533 PIN 369, Conf#: 9282

• 2013-07-16
• 2013-07-09
• 2013-07-02