Install on Docker
1. Clone the examples repository
Clone the aerospike-vector-search-examples repo to download a collection of configuration files and Docker Compose.
git clone https://github.com/aerospike/aerospike-vector-search-examples.git && \
cd aerospike-vector-search-examples/docker
2. Verify AVS feature-key
Ensure that your feature-key file features.conf is in ./config/ and includes the vector-service feature, as shown in the following example. If that service is not enabled, AVS will not start.
3. Start the Aerospike Database
You need a running Aerospike Database before you can use AVS.
docker run -d \
--name aerospike-cluster \
--network svc \
-p 3000-3003:3000-3003 \
-v ./config:/etc/aerospike aerospike/aerospike-server-enterprise:7.0.0.5 \
--config-file /etc/aerospike/aerospike.conf
The Aerospike Database must include a namespace for AVS internal system data storage and at least one namespace for vector and AVS index storage. See Configure Aerospike Database for AVS for more information.
4. Start AVS
When your Aerospike Database is up and running, you can start AVS.
docker run -d \
--name aerospike-vector-search \
--network svc \
-p 5000:5000 \
-p 5040:5040 \
-v ./config:/etc/aerospike-vector-search \
aerospike/aerospike-vector-search:1.0.0
You can now connect applications to AVS on port 5000.
Docker Compose
As an alternative to the steps above, you can use Docker Compose with all of the example applications. In any example application folder, simply run the docker-compose file to start the Aerospike Database, AVS, and the example application.
docker-compose up -d