remaining beta->prod task epic details

[fool]

most important

• get database storage moved out of docker so it persists between container restarts: move database storage onto persistent volume outside of docker, allowing r/w access from within docker #630
• automate new-environment setup: automate container post-start setup commands to whatever degree it is possible #631
• get in-container email working: get in-container email working #632
• get imagedir mapped correctly in nginx: figure out and implement a reasonable mapping for eventimages folder, adjusting nginx and docker to handle #633
• ensure that node is run correctly including npm install: confirm node is run correctly (including npm install at least when needed), within the container #634
• abstract env vars / settings into separate file: abstract settings from codebase and replace with git-safe placeholders that don't match prod #635
• figure out hugo best practices and document: figure out + implement + document hugo best practices #636

before testing:

• get node app.js "service-ified" so it auto-starts and restarts on crash: get node app.js "service-ified" to some degree, so it auto-restarts on crash within prod container #638
• figure out best practices for docker: figure out docker container best practices #639
• create less privileged user for system to run as (rather than root): configure less-privileged user to run mysql, nginx, and node inside docker container #640
• review TODOs: Review/complete any other TODOS #641
• how to/ should we constrain the number of spawned nginx/mysql processes: ensure reasonable # of threads configured for mysql and nginx #642

before production

• configure backups on beta: configure backups for beta #643
• create playbooks for starting from scratch Create playbooks for starting from scratch #644
  • ensure our playbooks for starting from scratch work even in prod (including migrating to/from api.shift2bikes.org)? use initial-config playbook to do migration? #645
• ensure that new beta host "listens/responds" as beta.shift2bikes.org and api.shift2bikes.org: ensure new server "works" as api.shift2bikes.org (maybe after DNS change + letsencrypt run) #646
• on-server credential review: review auth details on the server / in git #647
• smoke test all the things including via Aaron's client: smoke test  #648
• update netlify deploy process: update netlify deploy process #649
• remove legacy redirects: remove any legacy redirects that are no longer needed #650
• go live checklist in ZZProduction go-live checklist #637 , includes the following:
  • import legacy data
  • announce maintenance
  • announce new service and solicit feedback

and then:

• create "seed"/dummy database/image backups to enable local dev: create "empty"/dummy database/image backup and implement conditional restore #651
• document netlify details like [skip netlify] or maybe create an ignore script to not deploy if no changes to code: document netlify deploy details #652
• set up whatever monitoring we want: setup prod availability monitoring for new server #653
• remove remaining legacy stuff like docker-compose.yml, TODO etc: legacy detail cleanup #654
• update any relevant documentation around usage: better document end-state of the art for new devs #655

[fool]

All of the individual issues that are still relevant are still open, so I'll close this "epic"