Increase resilience for doing updates

[Joelius300]

At the moment, the receiver fails and loses 10 min of data if the database isn't up again after 10 minutes. Example: on 24.01.24 12:55Z it broke because it couldn't insert. The update process went for almost 20 minutes and the second time it managed to recover after 588 seconds so it was very close again. This ultimately led to 10 minutes lost data (which is irrelevant in the grand scheme but still).

1. Use docker compose profiles so that you can update the API etc. without having the database be down
2. Increase the time it can handle db outages, I think running into memory problems is still unlikely if 30 minutes worth of data is backed up (30 min data = 300 data points).
3. Lower the retry interval, 1 sec is too much. 5 suffices easily (still quicker than new data comes in)