Ah got it, yeah a queue of some sort is definitely useful when scaling up to buffer pre-inserted data. This is something on the OSS side we've kept open to implementation. However it's something that is highly coupled with infra footprint and internal SLA guarantees the user wants to preserve. It can range anywhere from just rely on client-side retries to setting up a HA Kafka cluster early in the ingestion pipeline.

Similar to Elastic - I think a lot of architectures are available to choose on that side when users want to scale.

Will reach out to connect!