Deploying Lambda-Spatial-DB

Getting Started

If you haven't already, please read through the database docs

1. Setup.

git clone https://github.com/geospatial-jeff/lambda-layer-spatial-db.git
cd lambda-layer-spatial-db
pip install -e .[dev]

2. Edit the database configuration in lambda_db/config.yml.

Variable	Description
db_name	Name of database. Also determines name of the Lambda Layer.
min_res	Minimum S2 cell resolution used when generating cell coverages.
max_res	Maximum S2 cell resolution used when generating cell coverages.
limit	Maximum number of S2 cells generated in a single cell coverage.
unique_id	Unique id field of database. This is used to filter back along the one-to-many relationship to prevent returning duplicate results.
compress	Determines if items are compressed (gzip) before inserted into database

You can create an optimized configuration for generating S2 cell coverages through the CLI. The following function suggests min_res, max_res, and limit parameters for the input geojson while optimizing for disk space (the algorithm is based on geometry-level statistics such as area, perimeter etc.). The command will also print out a variety of statistics about your data which can be used to inform your decision.

lambda-db analyze /path/to/feature_collection.geojson --optimize size

3. Build the database.

Once the database configuration has been updated you can build your database using the CLI:

lambda-db build /path/to/feature_collection.geojson

When the code finishes, your database will be saved in the lambda_db folder.

lambda_db/database.fs
lambda_db/database.fs.index
lambda_db/database.fs.lock
lambda_db/database.fs.tmp

4. Package the database as an AWS Lambda Layer and upload to AWS Lambda.

Once you've built the database you can generate an AWS Lambda Layer using the CLI:

lambda-db deploy -t <docker_tag>:latest --public

The above command will generate a lambda layer deployment package (lambda_db/lambda-layer.zip), upload the layer to AWS Lambda, and make the layer publicly available.

5. Check out your database info.

When finished, the lambda-db deploy function used in the previous step will print information about your database deployment. You can re-generate this information at any time using the lambda-db info CLI script.

6. Deploy Lambda Function (optional)

Once you have deployed the lambda layer to AWS Lambda your database is fully deployed and can be queried by attaching the Lambda Layer to a Lambda Function. The library provides a sample serverless deployment (lambda_db/serverless.yml) for testing out your database. Update the function's layer ARN in lambda_db/serverless.yml and deploy the service with sls deploy -v.

You can query your database by passing a geojson feature (with polygon geometry) to your lambda function's geoj argument.