Inspect a deployment in the operator console

Before you start

This operator workflow requires an issued runtime-console URL and environment access. A standalone public production-console address is not confirmed; the capture uses the real frontend with educational data.

Click path

1. Check environment and data-source indicators. демо-данные is not live workload state.
2. Open Развёртывания and search by agent name or deployment ID.
3. Inspect phase, available replicas, sessions and last wake.
4. Sleeping serverless workloads can normally have zero replicas. Use Пробудить only for an agreed manual wake; Усыпить changes runtime state.
5. Refresh after the action and wait for server readback. Inspect capacity, runtime trust and release evidence for failures.

Enlarge screenshot

1 2 3

1. Поиск развёртываний
2. Пробудить
3. Усыпить

Verify and recover

Only live server readback confirms a deployment change. Demo data does not prove execution. Capacity exhaustion and unsigned-image refusal have different remedies; report the run ID without service secrets.