At VerticalScope Inc., we use Vespa for powering our search and recommendation engine for 1000+ internet forums. We're providing a sample app here to sort of mimic how we use it internally using a sample forum dataset.
This is a sample implementation of Vespa search on Forum Data. We're going to be using data from https://diy.stackexchange.com/ for this example.
- Currently it's based on local Docker images. Kubernetes/Helm implementation coming up soon!
- Meanwhile, here are the docker instructions to get this up and running:
- Run a docker instance for MySQL using the command (We know this is unsafe for now. This will change when we implement Kubernetes yaml files):
docker run --name vespamysql -p 3306:3306 -p 33060:33060 -e MYSQL_ROOT_PASSWORD=<YOUR_PASSWORD> -e MYSQL_ROOT_HOST=% -d mysql:5.7
- Follow the instructions in the dataset/README.md to download and upload the https://diy.stackexchange.com/ to the MySQL docker instance
- Follow the instructions in the app/README.md to setup the Vespa docker on localhost
- Ensure that the above steps (dataset in SQL and Vespa docker setup) are complete and are running properly.
- Follow the instructions in the tools/vespaindexer/README.md to load all the forum data into Vespa docker.