Objective: Build and deploy faster and more reliably
 - Goals:
   - Decrease the amount of time needed from commit to deploy
   - Simple rollbacks
   - Automated testing
   - Legacy tool replacement
   - Distributable and/or centralized changelog access
   - Move off of legacy deployment infrastructure
 - Actions:
   - Continue development of courier
   - Continue development of skadi
   - Consolidate plethora of deployment solutions to courier
     - Audit
     - Config externalization?
     - Confirm service definitions
   - Explore different artifact creation (rocket, docker, etc)
   - Correct service definitions in puppet
   - Facilitate in providing management scripts for services
     - Post deploy execution
     - Service start/stop/restart with service

Objective: Enable developers to move faster
 - Goals:
   - New project to "prod" in 1 day
   - Dynamic environments
   - Standardized interfaces to logging and configuration
   - Support development on local machines (vagrant?, native?, etc)
   - Customizable backend services for testing
 - Actions:
   - Dev/Staging in EC2
   - Integrate with branchlab
   - Continue development on deploy tool
     - Provide feedback on deployment metrics and state
   - Provide standardized libraries for config and logging
   - Provide standardized template for build configuration
     - Continue manta development

Objective: Centralized logging (?)
  - Goals:
    - Setup or help setup centralized logging as a service (?)
    - Provide libraries for development teams to use for the service
    - Handoff service to operations
  - Actions:
   - Provide standardized libraries for logging
   - "Product-ize" the service for handoff to @ops
     - Documentation
     - Integration with deployment tools (?)
     - Standardized service management (start/stop/restart)

Objective: Support live configuration in k/v storage
  - Goals:
    - Provide facility to store k/v info for service configuration
    - Provide facility to update configuration as k/v updates
  - Actions:
    - Consul as k/v store (expanding current usage)
    - Consul-template
      - Explore....

Objective: More resiliant deployment system
  - Goals:
    - Eliminate one deploy blocking everyone else
    - Redudency
    - Ability to deploy when components of deployment system are down
  - Actions:
    - Chaos monkey
    - Audit existing architecture
    - Replace legacy components when possible
    - Monitoring + self healing ability of deployment tools

Objective: Datacenter management (mesos?)
  - Goals:
    - Integration with branchlab
    - Investigate best tool
    - Provide integration with existing tools
  - Actions:
    - Integrate branchlab with skadi, deploy, jenkins etc
    - Investigate

Objective: Cron as a service
  - Goals:
    - Distributed service to run crons
      - negates downtime of "cronhost"
  - Actions:
    - Identify solution
      - chronos?
    - Provide interface for developers to interact with cron service

Objective: Development environment orchestration
  - Goals:
    - Provide interface to bring up portions of infrastructure for integration testing
    - Ability to mix and match application revisions
  - Actions:
    - Explore solutions (atlas, cloud formation, fog + custom, etc)
      - Assume EC2 + private cloud via @ops

Objective: Create Twitch Coding Standards and Best Practices
  - Goals:
    - Provide language linters and standards docs
  - Actions:
    - Write standards docs and linters
    - Git hooks?

Objective: New merge tool
  - Goals:
    - Deprecate brigade
  - Actions:
    - Evaluate merging strategies across development teams
    - Provide migration strategy for teams wanting to move from legacy tools

Objective: Expand service discovery
  - Goals:
    - All services are discover-able
    - All services are discover-ed
  - Actions:
    - Continue consul integration for services

Objective: Libraries for healthchecking and test harness
  - Goals:
    - Standardized test harness ability for services
    - Integrate with service healthchecks
    - Multi-depth checks 
  - Actions:
    - Develop libraries that exercise multiple parts of system
      - synthetic transactions

Objective: Support for container-ized deploys
  - Goals:
    - Containers** as a first class deploy citizen
  - Actions:
    - Docker as a first class deploy citizen
    - Rocket?
    - etc

Objectives: Expand metrics collection within deployment system
  - Goals:
    - Metrics reported from deployment tools
  - Actions:
    - Instrument deployment tools
    - Dashboarding for metrics

Objectives: Expanded application testing
  - Goals:
    - Fuzz testing

Objectives: Inter-service authentication
  - Goals:
    - Provide libraries to integrate with oauth authentication
    - Application/Environment combos for keys

Objectives: Experiments
  - Goals:
    - Research and experiment with next gen technology
    - Provide thought leadership to developmnet organization

Objectives: Feature flags
  - Goals:
    - Work with development groups to implement feature flags for safer deploys
  - Actions:
    - Create standards for development teams to implement feature flagging
    - Dynamic config update and service restart

Objectives: Upgrade to Github Enterprise 2
  - Goals:
    - Update git.internal.justin.tv to GHE 2.x

Objectives: Send weekly web deploy metrics to QA
  - Goals:
    - Run weekly job that shows the number of branches deployed per
      deployment
