diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md index 8b3ffbb85..32662cce4 100644 --- a/docs/LOGBOOK.md +++ b/docs/LOGBOOK.md @@ -1,3 +1,18 @@ +## 2026-07-01 — 09:36 188 runner drain / 110 stale docker CPU metric correction + +**照主線修正的問題**: +- 為降低 110 Gitea / Harbor 高壓期間的輪詢與新 job 風險,188 已再次 drain:`awoooi-non110-runner.service`、`awoooi-non110-runner-keepalive.timer`、`awoooi-non110-runner-autostart.path` 讀回 `disabled / inactive`;`stockplatform-ubuntu-runner` 已 stop 並關閉 restart policy;未讀 `.runner` 內容。 +- 30 秒後 110 `node_load1/load5/load15` 仍高,Actions active container / process 仍為 `0`,代表 188 runner polling 不是唯一主因。 +- 110 `docker_container_cpu_cores{container_name="gitea"}` 讀值 `3.4019` 來自 stale `docker_stats.prom`:`node_textfile_mtime_seconds{file="/host/home/wooo/node_exporter_textfiles/docker_stats.prom"} 1.782778066e+09`,不可再把它當成最新 Gitea CPU truth;現行可信高壓證據是 node load / SSH timeout / Harbor 502 / current textfile timestamps。 +- 目前仍未恢復 110 control path:`ssh wooo@192.168.0.110` timeout,`registry.wooo.work/v2/` 與 `harbor.wooo.work/api/v2.0/health` 仍 502;下一步維持 110 local console / restored SSH 執行 `recover-110-control-path-and-harbor-local.sh --check`。 + +**驗證**: +- 188 drain readback:runner service / keepalive / autostart 均 `disabled / inactive`,`awoooi-cd-*` container 不存在。 +- 110 node exporter readback:`awoooi_host_gitea_actions_active_container_count=0`、`awoooi_host_gitea_actions_active_process_count=0`、`awoooi_host_load5_per_core` 仍高。 +- stale metric readback:`docker_stats.prom` mtime 與 `host_runaway_process.prom` / `awoooi_runner_failclosed_enforcer.prom` mtime 不一致。 + +**邊界**:本輪 runtime 寫入限於 188 user-level runner / idle stockplatform runner drain;未讀 secret / token / `.env` / raw sessions / SQLite / auth;未讀 `.runner` 內容;未使用 GitHub / `gh` / GitHub API;未 workflow_dispatch;未重啟主機、未 restart Docker / Nginx / K3s / DB / firewall。 + ## 2026-07-01 — 09:33 110 Harbor / Gitea pressure fail-fast gate **照主線修正的問題**: