Hybrid Search

This repo contains a public port of the hybrid search code used on aerospike.com. To run this locally, follow the steps below.

Note

Aerospike Vector Search (AVS) requires a feature key. Request one.

Pre-requisites

1. An Aerospike feature key with vector search enabled.
2. Docker & docker-compose

Running locally using docker-compose

1. Clone this repo and navigate to the hybrid-search directory.

   git clone https://github.com/aerospike-examples/hybrid-search.git && cd hybrid-search
   

2. Replace the config/aerospike/features.replace.conf and config/vector/features.replace.conf with a valid Aerospike feature key file.

   Note

   The feature key file must have a line item for vector-service

3. Create the containers by running:

   DOCKER_BUILDKIT=0 docker compose up -d

4. Load data into the database by running:

   docker exec -it -w /server search-server python3 load.py

   Note

   This will take some time. It's scraping and loading the Aerospike support knowledgebase.

5. Once the load script is finished, query the endpoint with query parameter q.
   For example:

   http://localhost:8080/rest/v1/search/?q=secondary index