NDT, Fluentd, Single-Volume Configs

This release includes three primary changes:

#234 to fix SIGEGV
#232 to reduce fluentd configs
#231 to use a single data volume for each experiment

New ndt-server and traceroute versions + cluster object/resource renames

This release includes these things:

• A fairly major rename of several key loadbalancer domain names, as well as renames of most all GCP resources. the renames mostly removed references to things like k8s-master and replaces them with platform-cluster. This is already running in production, but the bootstrap tooling has not be part of a release.
• A new version of traceroute caller, now v0.0.6.
• A new version of the ndt-server, now v0.10.0
• Addition of a new ndt experiment datatype named ndt5 to distinguish it from the new ndt7 type.
• A small fix to cache-control settings for the setup_k8s.sh script in GCS (don't ever cache it).

Update ndt-server version

This release publishes the v0.9.1 version of the ndt-server to production. This release should improve the test suite and s2c & c2s test metrics.

Brings production up to par with staging.

There have been 19 commits to fix various bugs in sandbox and staging since the last release. This release brings production up to date.

Small bugfix release.

Details for this release can be found in PR #217.

Version v1.0.0!

There are 486 commits to master since the previous release. We haven't been releasing new versions of this repository because until just a week or so ago the platform cluster didn't exist in M-Lab production space. It does now, so I expect new version will be forthcoming with more frequency now.

First Release of k8s_token_server

Merge pull request #37 from m-lab/sandbox-soltesz

Add Dockerfile to build token_server container