The scenarios open WebSocket sessions, subscribe to a topic, publish messages, and check that every subscriber sees the same events. There is also a reconnect scenario and a k8s manual-cluster variant that restarts a node mid-run.

How TF runs this

Each example follows the same pattern:

TF starts a small deployment of pubsub nodes
a workload opens WebSocket sessions and drives publish/subscribe behavior
an expectation checks that the nodes end up with the same topic state

Scenarios

basic_ws_roundtrip runs a local roundtrip check
basic_ws_reconnect runs a local reconnect scenario
compose_ws_roundtrip and compose_ws_reconnect run the same checks in Docker Compose
k8s_ws_roundtrip runs the roundtrip scenario on Kubernetes
k8s_manual_ws_roundtrip starts the nodes through the k8s manual cluster API, restarts one node, and checks that the topic state converges again

Run locally

cargo run -p pubsub-examples --bin basic_ws_roundtrip
cargo run -p pubsub-examples --bin basic_ws_reconnect

Run with Docker Compose

cargo run -p pubsub-examples --bin compose_ws_roundtrip
cargo run -p pubsub-examples --bin compose_ws_reconnect

Set PUBSUB_IMAGE to override the default compose image tag.

Run with Kubernetes

docker build -t pubsub-node:local -f examples/pubsub/Dockerfile .
cargo run -p pubsub-examples --bin k8s_ws_roundtrip

Prerequisites:

kubectl configured with a reachable cluster
helm installed

Optional image override:

PUBSUB_K8S_IMAGE (falls back to PUBSUB_IMAGE, then pubsub-node:local)

Run with Kubernetes manual cluster

docker build -t pubsub-node:local -f examples/pubsub/Dockerfile .
cargo run -p pubsub-examples --bin k8s_manual_ws_roundtrip