Photon Realtime 4.0: Adjusting per‑room message rate to mitigate concurrency latency

Goal

Determine how the documented 500 messages/second per‑room guideline in Photon Realtime 4.0 influences latency when many peers send events concurrently, and identify whether this behavior is consistent across Photon Cloud and self‑hosted Server deployments.

Constraints & Uncertainty

• The 500 msg/s figure is documented as guidance; the actual server may queue or drop messages depending on tier, region, and configuration.
• Latency spikes may arise from server‑side buffering or client‑side receive‑queue buildup, making it hard to isolate the root cause.
• Photon Cloud and self‑hosted Server expose different configuration knobs (e.g., message‑limit, fiber scheduling) that can shift the threshold.
• EventCaching and interest groups can alter traffic patterns for late‑joining peers, potentially creating transient spikes unrelated to steady‑state load.

Unresolved Questions

1. What is the exact per‑room message‑rate limit documented for Photon Realtime 4.0, and how does exceeding it affect latency when many peers send concurrently?
2. Are there configurable parameters in Photon Cloud or a self‑hosted Server that adjust this threshold, and what impact do those adjustments have on latency and disconnect behavior?