Deployment process for the new storm raster tiles

[eatyourgreens]

1. From the etl directory, run python/scripts/storm_layers_to_csv.py, to generate storm_layers.csv (you can skip this step if the file names and parameters haven't changed.)
2. Copy the new TIFFs into etl/raw_data/processed_data/STORM_RP_rasters/.
3. Run snakemake (you might need to comment out lines in the all workflow that load hazards and hotspots) to generate tiles in tileserver/raster/data.
4. Run the terracotta tile ingester from the root directory of the repo:

docker compose -f docker-compose.dev.yml up raster-tile-ingester

5. Copy the storm tiles to AWS:

scp -r ./tileserver/raster/data/storm* [email protected]:/var/www/tileserver/raster/data/

6. Copy the terracotta database to AWS:

scp -r ./tileserver/raster/data/terracotta.sqlite [email protected]:/var/www/tileserver/raster/data/

7. SSH to jamaica.infrastructureresilience.org, pull the latest frontend Docker image and restart the frontend service:

  cd /var/www
  docker pull ghcr.io/nismod/jsrat-frontend:canary
  docker compose -f docker-compose.stage.yml up -d frontend