Commit Graph

4 Commits

Author SHA1 Message Date
OG T
99be215e83 fix(monitoring): R1 Review 修正 — Blackbox DNS/PSA label/告警閾值
Critical: Blackbox Exporter replacement 從 K8s DNS 改為主機 IP (192.168.0.188:9115)
Important: Descheduler namespace 顯式宣告 PSA restricted labels
Suggestion: failedJobsHistoryLimit 3→1, 新增 MinioDiskUsageCritical 5% 告警

R1 Review by: 首席架構師 (Phase O-1)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-02 14:02:50 +08:00
OG T
6ce82ff883 fix(k3s): Phase O-1 基礎設施修復 — Descheduler + MinIO/Kali 監控
O-1.1: Descheduler securityContext 修復 (PodSecurity restricted 合規)
  - 新增 pod securityContext (runAsNonRoot, runAsUser:65534, seccompProfile)
  - 新增 container securityContext (allowPrivilegeEscalation:false, drop ALL)
  - 補齊 RBAC: namespaces + replicasets list 權限
  - 已部署驗證: CronJob 成功執行 (Status: Completed)

O-1.3: MinIO Prometheus scrape 配置 + 告警規則
O-1.4: Kali Blackbox TCP probe + 告警規則
  - MinioDown, MinioDiskUsageHigh, MinioOfflineDisk
  - KaliScannerDown

待手動部署: Prometheus config → .188, kubectl kubeconfig → 120/121

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-02 13:55:26 +08:00
OG T
0b68352fc2 feat(k3s): P2/P3 改進 - kube-state-metrics + Kured 時區修復 + Descheduler 調整
P2 改進:
- 新增 kube-state-metrics v2.10.1 (NodePort:30888)
- 新增 7 條 kube-state-metrics 告警規則 (NPD 整合)

P3 改進:
- 修復 Kured 維護窗口時區 (18:00→02:00 台北時間)
- Descheduler threshold 20%→30% (避免過度遷移)

首席架構師審查建議執行項目

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-03-28 22:23:42 +08:00
OG T
541565de48 feat(k4.2): Descheduler for pod rebalancing
- Deploy Descheduler v0.30.1 as CronJob
- Schedule: Every 2 hours
- Policies enabled:
  - LowNodeUtilization: rebalance when node < 20% usage
  - RemoveDuplicates: spread replicas across nodes
  - RemovePodsViolatingNodeAffinity: enforce affinity rules

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-03-28 22:03:54 +08:00