Configuration for a LangStream streaming instance.
Please note that support for Pravega.io is still experimental and is not recommended for production use.
Connect to a Pravega broker with the following configuration:
- The Controller URI
- The Scope
In Pravega, a scope is like a tenant, which is a namespace for your topics. In Pravega the topics are called "streams".
Configuration for a Pravega.io standalone instance looks like this:
streamingCluster:
type: "pravega"
configuration:
client:
controller-uri: "tcp://localhost:9090"
scope: "langstream"
Start the Pravega standalone instance with this command:
docker run -it --rm -p 9090:9090 -e HOST_IP=host.docker.internal -p 12345:12345 pravega/pravega:0.13.0 standalone
Start the LangStream docker chatbot example with this command:
export OPEN_AI_ACCESS_KEY=your-key-here
langstream docker run test \
-app https://github.com/LangStream/langstream/blob/main/examples/applications/openai-completions \
-s https://github.com/LangStream/langstream/blob/main/examples/secrets/secrets.yaml \
-i https://github.com/LangStream/langstream/blob/main/examples/instances/pravega-docker.yaml