Decide where the work belongs
Rank the critical action under burst
Name the request or workflow that becomes unusable first and separate read delay from failed writes or decisions. Include whether the system communicates overload clearly and whether people retry. A burst that blocks a durable action deserves priority even if a dashboard shows acceptable average throughput.
Measure frequency and concurrency reach
Group events by route mix, arrival rate, concurrency, request size, duration, and environment. Compare confirmed bursts with planned or hypothetical load. A frequent modest burst can outrank a rare extreme, while an unverified capacity projection should remain an evidence task.
Choose a bounded capacity slice
Select immediate investigation, a targeted queue or resource improvement, or collection of one missing controlled comparison. State the evidence that changes priority, such as a repeatable saturation boundary or recovery backlog. Keep broad capacity redesign separate from the first bottleneck that harms users.
What to carry forward
Produce a queue decision tied to critical work, confirmed burst reach, error behavior, and recovery debt. Include uncertainty about untested concurrency. The result is actionable when the next engineer has a bounded load shape to reproduce and a clear reason it outranks ordinary performance work.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow