We ran into some challenges with async inserts at highlight.io [2]. Namely, ClickHouse Cloud has an async flush size configured (that can't be changed AFAIK) that isn't large enough for our scale. Once you async insert more than can be flushed, you get back pressure on your application waiting to write while Clickhouse flushes the queue. We found that implementing our own batched flushing via kafka [3] is far more performant, allowing us to insert 500k+ RPS on the smallest cloud instance type.
Generally, any sort of async/batch inserts will get you decently far, but still will have limitations well before you get to million rows a second, mostly because it is really difficult to get your batch size large enough from individual producers without some sort of aggregation, which that aggregation is a challenge if you care about durability.
So often that means you need something like a Kafka to get the bulk ingest to really perform to get batch sizes large enough.
That kind of gets into one of the challenges of OSS observabilility systems, you don't want to make the dependencies insane for someone who only has a few thousand logs a second, but generally at some point of scale you do need more.
