From 44af0564b1bec2261b32b83e44ce7dac2074c07f Mon Sep 17 00:00:00 2001 From: pikastech Date: Sat, 18 Jul 2026 10:30:46 +0200 Subject: [PATCH] =?UTF-8?q?docs:=20=E8=B7=9F=E8=B8=AA=20NC01=20=E5=86=85?= =?UTF-8?q?=E5=AD=98=E6=81=A2=E5=A4=8D?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- docs/MDTODO/platform-resource-governance.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/docs/MDTODO/platform-resource-governance.md b/docs/MDTODO/platform-resource-governance.md index b5045127..606af098 100644 --- a/docs/MDTODO/platform-resource-governance.md +++ b/docs/MDTODO/platform-resource-governance.md @@ -13,3 +13,7 @@ ### R1.1 [completed] 执行 [UniDesk #2458](https://github.com/pikasTech/unidesk/issues/2458)：修复 NC01 上 HWLAB PaC/Tekton 大型内联 PipelineRun spec 被复制到 status.pipelineSpec、持续以约 1 MB 对象写入 Kine/SQLite并在 compaction 窗口阻塞 node lease，最终导致 k3s NodeNotReady 和 UniDesk backend-core endpoint 中断；将大段脚本拆入原生文件或复用 Task，补齐对象体积、Kine/WAL、API 延迟和 retention 的受控可见性，经正常 PR 自动交付并用 NC01 原入口验收，不以人工重启 k3s 或强删 Pod 作为最终修复，完成任务后将详细报告写入[任务报告](./details/platform-resource-governance/R1.1_Task_Report.md)。 + +### R1.2 [in_progress] + +继续执行 [UniDesk #1710](https://github.com/pikasTech/unidesk/issues/1710) 的 NC01 内存恢复：先用一次全局 `memory-distribution` 从业务、进程、cgroup、swap、僵尸进程和 Kubernetes 对象分布归因，目标是让 `/proc/meminfo` 的真实 `MemAvailable` 稳定超过 4 GiB；只执行 YAML 明确允许且 plan 判定可运行的受控 GC，禁止把 swap 计入目标、禁止触碰 Webterm/PK01、禁止重启 k3s或裸删 workload，若低风险候选不足则报告业务优化或废弃建议并收敛长期策略，完成任务后将详细报告写入[任务报告](./details/platform-resource-governance/R1.2_Task_Report.md)。