E0d/upgrade rabbit rebase

We're stopping rabbit, but it's part of the configuration the service,
so I didn't use manage:.  We may need to re-discuss the tags (mongo_3_0
works similarly when stopping mongo to manage config).

Files for building a docker container for rabbitmq

All apt-get over https

New third-party repos

Enforce correct rabbit version installed

Flag rabbit upgrading

Adds the CHECK_RABBIT_VERSION flag (true by default). If true and the
rabbitmq version found differs from the version specified in
RABBITMQ_VERSION, the role fails. If false in the same scenario, the
role continues but does not install rabbitmq (because it's already
installed).

This enables us to continue using this role without forcing us to
upgrade rabbitmq.

Rebase miss

Missed merge conflict marker

syntax and whitespace

latest rabbit.

Run edxapp on Docker for Devstack

Updated redis conf to work correctly (start/restart) with 16.04

fix supervisor configuration file path

Dockerfile for notifier on 16.04

Disable app preload in gunicorn for LMS & Studio.

Since the earliest days of edX, we've run gunicorn with preload_app as
True. This was primarily because the first handful of courses were using
the XMLModuleStore, which was extremely slow to initialize because it
had to read entire courses from disk and load them into memory at
startup.

XMLModuleStore courses stopped being created with the introduction of
Studio, and we've long since removed them from edx.org. Note: This
should not be confused with importing courses that are authored in XML.
XMLModuleStore courses are courses who use the disk instead of Mongo
for content persistence, and require a server restart for content edits
to show.

This commit sets things back to how Django is more commonly deployed
with gunicorn. It will make server startup time somewhat more expensive
because each worker process does init work, but it should make it
easier to safely enable features like database connection pooling
without having to worry about shared resources.

[PERF-305]

Quiet an ansible deprecation warnings

[DEPRECATION WARNING]: Instead of sudo/sudo_user, use become/become_user and
make sure become_method is 'sudo' (default).
This feature will be removed in a
future release. Deprecation warnings can be disabled by setting
deprecation_warnings=False in ansible.cfg.

Apply changes from the previous task.

Rabbit will only use 40% of RAM, allow this to go up

Just causes confusion with notifier.yml

Without this, resolv.conf won't be updated in our environment until the
lease expires and dhclient reruns.  The -n should prevent it from
*actually* changing interfaces, and instead just updates the conf files.

The linked documentation comes out ao .7 for a t2.large
We can also prevent rabbit from going to disk until more memory
is full, but that isn't needed right now.

Attempt to give the version tag attached by the pipeline a better name

LEARNER-406

infrastructure for building an AMI for a loadtest driver jenkins worker

The CD infrastructure shouldn't clobber the tag added by the underlyi…

* Add a new ansible role "loadtest_driver" for setting up any instance
  for running load tests.
* Make the existing "locust" role depend on loadtest_driver and pull out
  the redundant tasks.  The "locust" role is outside of the AMI-building
  code path, but we do use it for something else and want to keep common
  tasks shared.
* Make the "jenkins_worker" role depend on the new "loadtest_driver"
  role.
* Create a new ansible playbook for setting up loadtest drivers.
* create a new packer script to build an AMI for loadtest drivers.

matrix-auth was being overridden by Jenkins (down to 1.1)

https://github.com/jenkinsci/job-dsl-plugin/wiki#release-notes

We don't use any of the plugins impacted by the migrations from 1.50
to 1.52, but performing some manual testing of jobs.
https://github.com/jenkinsci/job-dsl-plugin/wiki/Migration#migrating-to-152

This appears to be because it wasn't pinned and the shipped version
of the plugin won.  We have pinning code in jenkins_master that we
can use, but we have to copy down that list to add ours (ansible
doesn't allow "recursive references" of a variable passed back to
itself.

Continue process of trying to group plugins and their deps.

Update sandbox base AMI.