Zero-downtime deployments

Handling zero downtime deployments:

In cases where multiple instances of a container are running, containers should be updated sequentially, monitoring the health check of a container and moving to update the next when the previous goes green

Context: https://www.reddit.com/r/node/comments/kupd8k/argus_a_typescript_based_successor_to_watchtower/