Lab: what happens when Kafka is down for an hour?¶

One script. PostgreSQL 17 and Kafka run in containers the script starts; "Kafka is down" is the container paused, so the broker neither answers nor refuses.

uv venv --python 3.13 .venv
uv pip install --python .venv/bin/python "omni-box[postgres,kafka]==0.2.1" "testcontainers[kafka,postgres]" asyncpg
.venv/bin/python kafka_down_lab.py

Twenty events go into the outbox, three more are published straight from a request path with a producer that was connected before the outage, Kafka is paused, the relay runs seven cycles, Kafka comes back, and the script prints the outbox statuses and the topic's message count at each step. The whole run takes a couple of minutes, most of it the paused cycles.