Understanding the SimpleL7Proxy — FAQ

Detailed questions and answers for each Quick Topic in the proxy discovery path. Each topic below has its own page.

FAQ Topics

How do user profiles determine when requests run? Where a request's priority comes from, when it takes effect, model overrides, and defaults.	What does a request's priority level control? How priority is structured (`PriorityKeys`/`PriorityValues`), what it controls in the proxy vs. APIM, and its effect on backends and retries.
How does the proxy keep queueing fair across users? Per-priority worker pools and the `UserPriorityThreshold` mechanism that stops one user monopolizing a level.	How does APIM determine which backends receive requests? Priority-aware backend selection, throttling/`429` handling, `LoadBalanceMode` selection, failover, and requeue behavior.
When should traffic go directly to a backend or through APIM? What each mode is, when to use direct mode, and when to route through APIM.	How does the proxy stay resilient and autoscale with ACA? Backpressure, circuit breaking, KEDA scale triggers, health probes, replica lifecycle, and draining.
When should clients stop waiting synchronously? What async solves, how a request enters async mode, the trigger timeout/`202` flow, and when to stay synchronous.	How does the proxy handle observability and telemetry? Telemetry channels (App Insights, `ProxyEvent`/`EVENT_LOGGERS`), event fan-out, event contents, streaming token counting, and custom sinks.