Deployment tooling/Cabal/2016-09-19

Next release/blocking

Goal early next week end of this week—ops has an offsite next week

•  Done Allow per-environment config-override
• Local config deploys should use the target's current version
  • TODO: Break out proposed solution to separate task
  • TODO: Store DEPLOY_HEAD on target
• Scap config management: Jinja2 fills templates with Pythonic values
  • Workaround available, but a nasty one
  • May not make next release
  • Should be high priority

Task prioritization

Factors to consider:

• Highest priority: data-loss/security issues/blocking extended use
• Incorrect functionality vs. Enhancement
• Workaround available vs. None
• Frequency of occurance (affects all scap users vs. affects some)

Tasks to prioritize

• Scap3 submodule space issues (all users, no workaround, enhancement)
  • ORES uses 5 GB IIRC
  • SCB shares space with lots of other services
  • Stored with logs
• scap to reload a service instead of restart (all users, command workaround, enhancement)
• Scap3 config references to deployed directory (all users, no workaround, incorrect/non-obvious functionality)
• puppet should run deploy-init command once after cloning the deploy repo (already done, just need full path))

As Always

• Phase 2
• Phase 1
• Workboard https://phabricator.wikimedia.org/tag/scap3/
• Etherpad backed up to https://www.mediawiki.org/wiki/Deployment_tooling/Cabal
• Future document https://www.mediawiki.org/wiki/Deployment_tooling/Future
• Spreadsheet: https://docs.google.com/spreadsheets/d/1MlEsFxrLvdZdV_G82WEAIvBXr7ArO7nCEKaFClHhJEw/edit#gid=0