# Capacity And Resource Pressure

## Question

Is a capacity or quota boundary directly throttling the affected service, or is
high utilization only a correlated background condition?

## Enter When

Use for sustained CPU/memory/disk/region pressure, quota saturation,
resource-control queueing, throttling, autoscaling lag, or capacity alerts.

## First Batch

1. Align the service result with the exact resource/queue/throttle signal on
   the same entity and samples.
2. Compare usage with the real limit/quota and one pre-incident or peer
   baseline.
3. Check demand growth versus an internal backlog/inefficiency alternative.

Use same-entity range trends for this capacity question. Workload identity and
per-execution attribution are not part of this Scenario's default objective;
they belong to a separately selected Scenario. Do not widen the objective to
workload detail merely because utilization increased. Include bounded
configuration, scaling, maintenance, or deployment change correlation only
when it could materially change the capacity conclusion.

## Decision

High utilization is not a mechanism without saturation, queue, throttle, or
service degradation. Preserve units and distinguish requested/configured quota
from effective runtime capacity. For resource control, queue time can remain
high despite apparently sufficient aggregate quota when workload distribution
is uneven.

## Strongest Alternative And Stop

Check one direct wait/queue path or workload-volume measure. Stop after
capacity explains the result or is ruled out with actual headroom. Never scale,
change quotas, or modify resource groups automatically.