-
Notifications
You must be signed in to change notification settings - Fork 476
Description
In the context of Azure Functions, a cold start scenario can involve multiple concurrent requests. Typically, one request triggers the initialization of the runtime (and is tagged as a cold start), while other concurrent requests wait for the runtime to be ready. Once initialization completes, all waiting requests are processed.
While only the first request is tagged as a cold start, the others are also impacted by the cold start—they experience increased latency due to waiting for the runtime to initialize.
Additionally, there can be more requests arriving while the runtime is still getting ready, and all of these requests will also wait and be affected by the cold start.
If I'm analyzing response times, I’d want to identify all requests affected by the cold start, not just the one that triggered it. This would help in understanding the root cause of latency spikes more accurately.