Incident response and capstones · Lesson 9 of 12 · 6 min

Prepare a model rollback

A model rollback restores a compatible, previously verified release tuple. Preserve the failing candidate evidence so recovery does not erase the investigation.

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-incidents-capstones-009.json. Separate the recorded facts from your assumptions.
Hint & worked example
A model rollback restores a compatible, previously verified release tuple. Preserve the failing candidate evidence so recovery does not erase the investigation.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-incidents-capstones-009.json
```
Common mistake to avoid: Mix the old model with an arbitrary runtime and delete the failure logs.
2
Isolate the deciding signal
Select the pair_verified field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Restore the verified m3/r2 tuple while retaining candidate evidence.
A successful JSON query isolates pair_verified.
```
jq .pair_verified /opt/ai-lab/cases/aiops-incidents-capstones-009.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: restore-verified-tuple, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-incidents-capstones-009-decision.txt.
Hint & worked example
Restore the verified m3/r2 tuple while retaining candidate evidence.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'restore-verified-tuple' > /home/learner/lab/aiops-incidents-capstones-009-decision.txt
```
Common mistake to avoid: Mix the old model with an arbitrary runtime and delete the failure logs.

Understand the pattern

A model rollback restores a compatible, previously verified release tuple. Preserve the failing candidate evidence so recovery does not erase the investigation.

Interpret the evidence

Restore the verified m3/r2 tuple while retaining candidate evidence.

A tempting mistake

Mix the old model with an arbitrary runtime and delete the failure logs.

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-incidents-capstones-009.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "rollback_model": "m3",
  "rollback_runtime": "r2",
  "pair_verified": true,
  "candidate_evidence_saved": true
}

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

$ jq .pair_verified /opt/ai-lab/cases/aiops-incidents-capstones-009.json

true

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

$ printf '%s\n' 'restore-verified-tuple' > /home/learner/lab/aiops-incidents-capstones-009-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 “restore-verified-tuple” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Prepare a model rollback” justify?Make the change now and gather the missing evidence afterward.Restore the verified m3/r2 tuple while retaining candidate evidence.Mix the old model with an arbitrary runtime and delete the failure logs.

Apply it elsewhere

A colleague concludes: “Mix the old model with an arbitrary runtime and delete the failure logs.” What should you do?Reject that conclusion; return to the case evidence and its stated limits.Accept it because the training command completed.Treat this fixture as a live measurement.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Build a complete evidence bundle 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