Shutdown staging

[BryanQuigley]

1. Confirming how any needed development will continue without a staging site (and confirming we are going with fully taking down staging vs just turning off ECS after use or something similar).
2. Shutting down staging via terraform destroy (and double checking if could be used by both)
3. Checking for other items that can be removed that weren't in terraform
4. Documenting how to do development going forward.

[BryanQuigley]

This should wait until we've finished other issues that are easier to do with staging, namely performance testing, bugs and maybe react upgrade.

[KlaasH]

Just to note here: this hasn't been done yet, but we've moved in that direction.

• The desired count for the EC2 auto-scaling group is set to zero, so there's nothing running there.
• The ECS cluster and service are still there and still have the desired task count set to 1, but since there aren't any instances in the cluster, they can't run. I don't think it costs money for them to just exist, though I would rate my confidence on that low, since AWS does like to charge for the mere existence of things in unanticipated ways.
• The database is now a t3.micro, so I think it only costs about $15/month to keep that and its storage active. Since it would be a bit more of a thing to spin that up if we took it down completely, it might make sense to leave it unless we're confident we won't need it for a long time.

[dmcglone]

@KlaasH, should we consider this issue closed now? That resolution sounds good to me.