Observability and tracing · Lesson 4 of 15 · 7 min

Compare throughput signals

Requests per second and tokens per second measure different throughput. Workload length can change token throughput without the same change in request throughput.

Open the lab Course map

Mastery 0/3Run itExplain itApply it

learner@clairos:~ $

↑ history · Tab complete · Ctrl+C cancel · Esc, Tab to leaveExit 0

Your objectives

Hints are always available. Wrong turns never cost XP.

1
Read the case evidence
Read /opt/ai-lab/cases/aiops-observability-tracing-004.json. Separate the recorded facts from your assumptions.
Hint & worked example
Requests per second and tokens per second measure different throughput. Workload length can change token throughput without the same change in request throughput.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-observability-tracing-004.json
```
Common mistake to avoid: Treat requests per second and tokens per second as interchangeable.
2
Isolate the deciding signal
Select the mean_output_tokens field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Keep request and token throughput separate; the observed mean is 30 output tokens per request.
A successful JSON query isolates mean_output_tokens.
```
jq .mean_output_tokens /opt/ai-lab/cases/aiops-observability-tracing-004.json
```
Common mistake to avoid: Using a different case or an unrelated signal as the justification.
3
Record your operational decision
Choose the justified code: keep-throughput-units, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-observability-tracing-004-decision.txt.
Hint & worked example
Keep request and token throughput separate; the observed mean is 30 output tokens per request.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'keep-throughput-units' > /home/learner/lab/aiops-observability-tracing-004-decision.txt
```
Common mistake to avoid: Treat requests per second and tokens per second as interchangeable.

Understand the pattern

Requests per second and tokens per second measure different throughput. Workload length can change token throughput without the same change in request throughput.

Interpret the evidence

Keep request and token throughput separate; the observed mean is 30 output tokens per request.

A tempting mistake

Treat requests per second and tokens per second as interchangeable.

All names, measurements, prices, versions, and thresholds in this exercise are fictional training data. The evidence supports this bounded scenario, not a claim about a live provider.

Worked example and expected result

$ cat /opt/ai-lab/cases/aiops-observability-tracing-004.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "requests_per_second": 20,
  "output_tokens_per_second": 600,
  "mean_output_tokens": 30
}

Exit status: 0. The command completed successfully in the training fixture.

$ jq .mean_output_tokens /opt/ai-lab/cases/aiops-observability-tracing-004.json

Exit status: 0. The command completed successfully in the training fixture.

$ printf '%s\n' 'keep-throughput-units' > /home/learner/lab/aiops-observability-tracing-004-decision.txt

(no output; exit 0)

Exit status: 0. The command completed successfully in the training fixture.

Try a variation

Name the assumption that would have to change for “keep-throughput-units” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Compare throughput signals” justify?Treat requests per second and tokens per second as interchangeable.Make the change now and gather the missing evidence afterward.Keep request and token throughput separate; the observed mean is 30 output tokens per request.

Apply it elsewhere

A colleague concludes: “Treat requests per second and tokens per second as interchangeable.” What should you do?Treat this fixture as a live measurement.Reject that conclusion; return to the case evidence and its stated limits.Accept it because the training command completed.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Inspect error-rate evidence Try an incident drill

Anonymous lesson and mastery progress lasts for this session. Sign in to keep synced lesson XP between sessions; mastery stays device-local. No speed bonus, lost lives, or hint penalties.

Keep learning

Case methodology: evidence, limits, decision