Spin up GCP cloud services to run scraping scripts

[emigre459]

The theory is that we'll be able to run our scripts faster on a dedicated server with high-speed networking...we shall see!

• [ ] Create evlens GCP organization node
• [x] Parallelize location ID scraping from dev map
• [ ] Update scripts/scrape_location_ids.py to use parallelized code and see if you can beat 4 seconds per search location (when no pins found)
• [x] Setup scraping code to save results to BigQuery
• [x] OPTIONAL: Push hexagon locations from h3_hexagons_usa.pkl (and IDs from index) data we use into BigQuery for hopefully faster referencing (e.g. using an OFFSET call per scraping worker)
• [x] Save error screenshots to Cloud Storage
• [x] Build out Dockerfile
• [x] Build and push docker image

[emigre459]

Can't figure out why, but found that containerizing the selenium code causes it to hang indefinitely after it exits the login dialog on Plugshare. Abandoning the container approach for a messier "just run it on a GCP VM after installing what you need" approach to get things moving.

[emigre459]

VM spun up