ParentFull threadx_foo_x·Latency was a concern at launch but it has improved a lot, I see less than 20 ms these days. Depends on scale though, we have moderate sustained load 24/7View on HN