Observe a Blocked Worker from a Separate Control Path

These examples and illustrative results are independently authored teaching materials, not measured model results.

Use case

W1 waits on C5, so an in-queue status command also waits. Teaching snapshot at UTC10:05 contains C5 at 10:00 with no matching result in inspected scope. That is scoped observation, not proof of global deadlock.

Mechanism

Read bounded current traces/status from an independent path, verifying session/time/coverage. Correlate calls/results and distinguish stale, truncated or denied evidence. Reading differs from stop/steer/restart capability.

Bad example

Queue inspection behind W1 and claim monitoring ran; infer deadlock/restart from an old unmatched log.

Good example

Independently read the authorized W1 snapshot at UTC10:05. Report C5 at 10:00 unmatched within that scope, observed age five minutes, with completeness/current identity checked. Missing/truncated telemetry stays unknown, not healthy-empty. Do not claim control or termination; further actions retain scope requirements.

Why the change matters

In-queue monitoring depends on recovery. Separate observation provides evidence while coverage/freshness prevent gaps becoming runtime facts.

Observable expectation

The fixture returns a snapshot or read failure without W1 dequeuing. Distinguish scoped unmatched, global uncertainty and time.

Uninspected ranges may contain results; snapshot analysis does not establish actual timeout cancellation. No private sessions are read.

Limits

Telemetry delay can explain stale logs. Independent reads still have permission/freshness limits; source CLI/paths grant no cancellation guarantee.