[guide] docs: add fleet-health.md — health verdicts and remediation hints (#5577) - #5609
[guide] docs: add fleet-health.md — health verdicts and remediation hints (#5577)#5609kubestellar-hive[bot] wants to merge 1 commit into
Conversation
…5577) The #5577 feature (b35cb9c, 7157d30, 0ba5a9f) gave every hive a green/amber/red/unknown health verdict on /fleet with a WHY chip, signature-matched remediation hints, and detector states — but no operator documentation existed anywhere. Add src/docs/fleet-health.md covering the four verdict states, base precedence (App-broken > provider limit > budget > blocked agents > generic stale-output), the cause-to-remediation table from pkg/hub/remediation.go, the detector-state layering (error-streak re-explains generic reds at 3+ consecutive model-call failures; consent-wedge / no-cadence / channel-lag demote green only), and the L1/offline exemptions. Link it from the src/docs/README.md index and cross-link from health-checks.md. Fixes #5606 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: sec-check <sec-check@hive.kubestellar.io>
kubestellar-prow
Bot
added
the
dco-signoff: yes
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
kubestellar-prow
Bot
added
the
size/M
kubestellar-hive
Bot
added
documentation
Ported from the duplicate PR #5609: health-checks.md now points readers at the /fleet output verdict page, making the reachability-vs-output split discoverable from both directions. Signed-off-by: Andy Anderson <andy@clubanderson.com>
|
Closing as a duplicate: this and #5612 both added the fleet-health operator docs for #5606, and #5612 has now merged to v4. The merged page carries everything here plus the ACMM output-banding table, the full seven-step precedence (including the all-paused and repo-Issues-disabled cases), detector semantics verified against the source constants (error-streak grace/max-age, the one-hour consent-wedge window, the 100k budget floor), the nil-vs-empty carry-forward rule, and a symptom-to-fix troubleshooting table. The one thing this PR had that #5612 did not — the cross-link from health-checks.md back to the new page — has been ported in the follow-up #5615, with credit to this PR in the commit. Thanks for catching the gap; the coverage judgment was right, the other PR just got there with more depth. |
1 participant
Documentation Fix
Adds operator documentation for the #5577 fleet health verdict + remediation hints feature (merged in b35cb9c, 7157d30, 0ba5a9f), which previously had zero docs coverage.
Files/cluster claimed (docs only — no source code touched):
src/docs/fleet-health.md(new) — the four verdict states, base-verdict precedence (App-broken > provider limit > budget > blocked agents > generic stale-output), the cause→remediation table fromsrc/pkg/hub/remediation.go(app-broken, login-stuck, budget-exhausted/misconfigured, error-streak, consent-wedge, no-cadence, hold-stale, channel-lag), detector-state layering rules (error-streak re-explains a generic red at 3+ consecutive model-call failures; consent-wedge/no-cadence/channel-lag demote green only; green never carries a hint), and L1/offline exemptions.src/docs/README.md— index entry after "Dashboard route and health checks".src/docs/health-checks.md— one-line cross-link to the new page.Disjoint from open hold-gated PRs #5559 (UPGRADE.md) and #4032 (proxy source).
Fixes #5606
Filed by guide agent (ACMM L5 — hold-gated mode). Hold-gated: human review required.
— hive: agent=guide backend=copilot model=claude-sonnet-4-6