Closed cgolubi1 closed 4 months ago
This should be basically ready to go. I'll setup the PR and checkout tomorrow morning, in case anything changes between now and then for some reason.
Starting the prep work now.
My workspace is ready to go. Waiting until 16:00 UTC, since that's when i said i would kick this off.
Okay, off we go, taking the prod instance down now.
Disassociated EIP from prod instance.
Created NLB, waiting for it to finish provisioning.
Okay, done. Kicking off the deployment now.
That's running --- the next step takes about 15 minutes (it will always take that long, but usually the site is still available running the old container while it happens --- the downtime is only because of the cutover from EC2 to ECS).
Okay, the site appears to be running. Let's check things out.
I can take normal actions on the site, and when i do, log entries show up on the container. We received the e-mail with the container's IP address. There's not really much else to check; either it works or it doesn't, and it seems to work.
Closing this out. Let's hold off on merges for about a day just in case anything urgent comes up, and then go from there.
Deploy the changes made for #2908 and tested on staging on #2937 to the production site.
Plan (copied from steps i ran for staging on #2937):
buttonmen_ecs_config.json
is knowable. So the next step is to fully fill out all entries for the stage in that file.Ideally, that's it. If anything goes wrong, roll back and revisit.