Speeeed, (although current speed isn’t bad, it could help w collab/ot)

Not currently too pressing

usage graphs and metrics to know when and where the app is slow/error-ing

can be reverted from backup, regular backups?

Alert Email when usage/response time passes a threshold /percentile

record logging data, ideally interactively/searchably. configureable data retention period

can invite other ppl into the account to view metrics and logs? team debugging

Memory leaks shouldn’t bankrupt me

Switch to using set timeout for short time stuff

i dont want to set this up myself, and take the chance that the db doesn’t work irl when needed

Cron jobs: scheduler for tasks than run every day? Eg every Sunday

Upgrade Postgres version locally first as test

backup the db

create a follower db

bonus: webhooks for downtime/slowdown errors so i can make a simple api status page

local branch: update kinopio-server node version and deps

turn off the server

git: push branch when done

upgrade the follower db

try on dev first

railway: deploy from branch

railway: port env vars

on both machines

railway: test routes, and confirm heroku db connection works

netlify: re-route api.kinopio.club to use railway app

turn on the server

move the db over

this should restart the server

wait a week and see if anyone reports any issues with it

use fixed values for heroku db connection (same as postico)

railway: create new postgres db

test that the new db is taking the writes

kinopio-client: take client down for 15 mins (replace w temp 401 page), share w ppl a couple days in advance and on the day of

turn off old primary

railway: copy heroku db to railway db

Show client migration downtime warning notification on day of

railway: update railway db env values to point to new db

railway: test some GET routes

kinopio-client: revert to working state

🎉 announce the migration is complete