This message was deleted.

We plan to ingest via OTEL, right now we have flink jobs which compute metrics every 30 seconds and push it to a kafka topic. From there we have custom collector that converts it to prometheus format