Quick Start, using Docker

This guide shows how to install and run Vespa on a single machine using Docker - it deploys an application, feeds some data and issues queries. See Getting Started for troubleshooting, next steps and other guides.

This guide is tested with vespaengine/vespa:8.444.18 container image.

Running Vespa in Docker

1. Validate the environment:

   $ docker info | grep "Total Memory"
   or
   $ podman info | grep "memTotal"
   

   Make sure you see at minimum 4 GB.

2. Install the Vespa CLI:

   Using Homebrew:

   $ brew install vespa-cli
   

   You can also download Vespa CLI for Windows, Linux and macOS.

3. Set local target:

   $ vespa config set target local
   

4. Start a Vespa Docker container:

   $ docker run --detach --name vespa --hostname vespa-container \
     --publish 8080:8080 --publish 19071:19071 \
     vespaengine/vespa
   

   The port 8080 is published to make the search and feed interfaces accessible from outside the docker container, 19071 is the config server endpoint. Only one docker container named vespa can run at a time so change the name if needed. See Docker containers for more insights.

5. Initialize myapp/ to a copy of a sample application package:

   $ vespa clone album-recommendation myapp && cd myapp
   

6. Deploy it:

   $ vespa deploy --wait 300
   

7. Feed documents:

   $ vespa feed ext/documents.jsonl
   

8. Issue queries:

   $ vespa query "select * from music where album contains 'head'" \
     language=en-US
   

   $ vespa query "select * from music where true" \
     "ranking=rank_albums" \
     "input.query(user_profile)={pop:0.8,rock:0.2,jazz:0.1}"
   

   $ vespa query "select * from music where true" \
     "ranking=rank_albums" \
     "input.query(user_profile)={pop:0.8,rock:0.2,jazz:0.1}" \
     "presentation.format.tensors=short-value"
   

   This uses the Query API. Note that a query's language is a factor when doing text matching, refer to linguistics to learn more.

9. Get documents:

   $ vespa document get id:mynamespace:music::a-head-full-of-dreams
   

   $ vespa visit
   

   Get a document by ID, or export all documents - see /document/v1 and vespa visit.

Next steps

Check out getting started for more tutorials and use cases which Vespa is great for.

$ docker rm -f vespa