docs: move ops runtime tracking under docs
This commit is contained in:
126
docs/ops_center_runtime/HANDOFF_20260419_0107.md
Normal file
126
docs/ops_center_runtime/HANDOFF_20260419_0107.md
Normal file
@@ -0,0 +1,126 @@
|
||||
# HANDOFF 2026-04-19 01:07
|
||||
|
||||
## 当前最高优先级
|
||||
|
||||
唯一最高优先级:
|
||||
|
||||
- 执行 `C1-Acceptance Re-run`
|
||||
|
||||
原因:
|
||||
|
||||
- `B2-Journal Permission Closure` 在运行态上已经具备完成证据
|
||||
- 两台节点都已出现:
|
||||
- `logs.collect ok=True`
|
||||
- `domaincheck-node-agent` 重启成功
|
||||
- `registered: Agent 注册成功`
|
||||
- 当前只剩旧 acceptance 结果尚未被新运行态覆盖
|
||||
|
||||
## 本轮确认到的运行态证据
|
||||
|
||||
`mainland-controller-01`:
|
||||
|
||||
- 旧执行记录中已经出现:
|
||||
- `action=logs.collect ok=True`
|
||||
- 之后又完成:
|
||||
- `domaincheck-node-agent.service` 停止 / 启动
|
||||
- `starting node agent`
|
||||
- `registered: Agent 注册成功`
|
||||
|
||||
`mainland-worker-01`:
|
||||
|
||||
- 旧执行记录中已经出现:
|
||||
- `action=logs.collect ok=True`
|
||||
- 之后也完成:
|
||||
- `domaincheck-node-agent.service` 停止 / 启动
|
||||
- `starting node agent`
|
||||
- `registered: Agent 注册成功`
|
||||
|
||||
## 当前判断
|
||||
|
||||
当前可以做出的最小结论:
|
||||
|
||||
- 日志权限收口在运行面已经打通
|
||||
- Node Agent 当前在线且能重新注册
|
||||
- 不需要再回头排查 SSH、bootstrap、heartbeat、队列消费
|
||||
|
||||
当前还不能直接签收上线的唯一原因:
|
||||
|
||||
- 控制面上仍保留旧 acceptance run 的 `attention` 结果
|
||||
|
||||
## 下一轮若继续,第一步先做什么
|
||||
|
||||
下一轮第一步只做这一件事:
|
||||
|
||||
- 分别对 `mainland-controller-01` 和 `mainland-worker-01` 重跑 acceptance
|
||||
|
||||
然后依次做:
|
||||
|
||||
1. 记录新的 acceptance `run_code`
|
||||
2. 确认两条 run 是否全绿
|
||||
3. 复核:
|
||||
- `go-live-summary`
|
||||
- `stack-diagnosis`
|
||||
- `release-launchpad`
|
||||
4. 输出上线签收证据摘要
|
||||
|
||||
## 推荐模型与推理等级
|
||||
|
||||
继续推荐:
|
||||
|
||||
- `GPT-5.4`
|
||||
- `high`
|
||||
|
||||
原因:
|
||||
|
||||
- 当前还是高耦合收口阶段
|
||||
- 但已经不需要切超高推理
|
||||
|
||||
## 现在不要做的事情
|
||||
|
||||
不要做:
|
||||
|
||||
- 新功能
|
||||
- 新页面
|
||||
- 新模块
|
||||
- 控制面增强项
|
||||
- 新专题文档
|
||||
- 正式发布动作
|
||||
|
||||
## 给用户的下一任务卡片
|
||||
|
||||
按下面这张任务卡继续:
|
||||
|
||||
```text
|
||||
任务名:C1-Acceptance Re-run / 两台大陆节点验收重跑
|
||||
|
||||
目标:
|
||||
1. 用新的运行态覆盖旧 acceptance 失败结果
|
||||
2. 确认 mainland-controller-01 与 mainland-worker-01 是否都已通过标准接管验收
|
||||
3. 为上线签收准备最终证据
|
||||
|
||||
执行位置:
|
||||
- 海外控制面主机
|
||||
|
||||
执行命令:
|
||||
cd /www/wwwroot/getDomain
|
||||
bash domain-api/deploy/multi-region/drive_ops_center.sh node-acceptance-run http://127.0.0.1:8100 mainland-controller-01 cli
|
||||
bash domain-api/deploy/multi-region/drive_ops_center.sh node-acceptance-run http://127.0.0.1:8100 mainland-worker-01 cli
|
||||
|
||||
然后继续执行:
|
||||
curl -s http://127.0.0.1:8100/api/v1/ops/go-live-summary
|
||||
echo
|
||||
curl -s http://127.0.0.1:8100/api/v1/ops/stack-diagnosis
|
||||
echo
|
||||
curl -s http://127.0.0.1:8100/api/v1/ops/release-launchpad
|
||||
|
||||
回传给 Codex 的结果:
|
||||
1. controller 新的 acceptance run_code 和最终状态
|
||||
2. worker 新的 acceptance run_code 和最终状态
|
||||
3. go-live-summary 当前状态
|
||||
4. stack-diagnosis 当前状态
|
||||
5. release-launchpad 当前推荐动作
|
||||
```
|
||||
|
||||
## 一句话结论
|
||||
|
||||
当前项目已经越过“日志权限修复”这一步,下一步不需要再修环境,只需要把两台大陆节点的 acceptance 重跑一遍,用新结果完成上线前签收判断。
|
||||
Reference in New Issue
Block a user