Allow continuous deployment of euctr-tracker-data from GitHub #104
Description
We recently wanted a more automated rollback process after a failed scrape led to bad data on the production site. We reverted the new data in the euctr-tracker-data repo on GitHub, however it still did not go live.
The current deploy script pulls the data repo from GitHub every six hours. However, it only runs /deploy/restart-web-services.sh when it commits new data. This means that if we make a new commit to the euctr-tracker-data repository from elsewhere, it does not get fully deployed.
We could just run restart_web_services every time we do git pull, but that would be a bit excessive. A better solution might be to verify the output of running the initial git pull and restart only when we see new data from the Github origin.