Kubernetes GPU Guide 2026: Scheduling, Sharing, Monitoring
Run GPUs on Kubernetes in 2026: install the NVIDIA GPU Operator, share GPUs with MIG or time-slicing, serve LLMs with vLLM, and monitor with DCGM.
Topic
10 篇笔记
Run GPUs on Kubernetes in 2026: install the NVIDIA GPU Operator, share GPUs with MIG or time-slicing, serve LLMs with vLLM, and monitor with DCGM.
📝 摘要 基于 Proxmox VE 8.4.0(i440fx/SeaBIOS)部署 Windows 7 Professional SP1 x64 虚拟机的完整实录,重点解决 VirtIO 驱动安装的兼容性陷阱。核心策略是先用 IDE 模式装系统、再切换 VirtIO Block 控制器,并写入 CriticalDeviceDatabase 注册表条目以避免 蓝屏。 核心概念: Windows…
📝 摘要 本文汇总了 PaddleOCR-VL 服务在 vLLM 部署下的效果问题排查结论及完整部署方案。客户反馈 PaddleOCR-VL 丢失版面分析能力,根本原因是直接用 OpenAI SDK 裸调 vLLM 的 接口,只调用了 VLM 识别模型而绕过了版面分析模型 PP-DocLayoutV3。PaddleOCR-VL 的正确用法是两阶段 pipeline:版面分析 + VLM 识别。文…
Increasing throughput sometimes requires reducing parallelism. A look into why this is the case for Vitess + MySQL databases
📝 摘要 本文记录了 CSGHub 集群一次由 OpenEBS cStor 3.6.0 存储故障引发的 Redis Pod CrashLoopBackOff 排查全过程,故障波及四个 CStorPoolInstance 全部进入只读状态。排查揭示三层根因:池容量达 82% 触发 cStor 硬编码的 85% 只读保护阈值(不可配置、不可自动解除);清理容量后卷副本 rebuild 陷入状态机死锁…
Learn about the three optimization levers for distributed AI inference: prefill/decode disaggregation, KV cache strategy, and speculative decoding.
Stop drowning in Kubernetes dashboards. Learn how consistent signal architecture, not more tooling, is what actually cuts incident resolution time in half.
From SLIs to LLM monitoring: 9 practices that separate mature teams from dashboard collectorsMost teams have dashboards. Few have observabil...
这次问题的核心现象是: 最终定位: OpenEBS 后端已经扩容成功,但 Linux device-mapper(multipath)层没有刷新容量,导致文件系统仍看到旧大小。 二、问题现象 PVC describe 看到: 但 events 中持续出现: 以及: 同时又有: 这是一个非常“矛盾”的状态。 --- 三、第一次误判 最开始以为: 因为: - external-resizer 持续 r…
Open: Pasted image 20260513104709.png30-Resources/Clippings/2026/2026-05/assets/03a146e5af0e9e40f7dbae15116436eaMD5.jpg Kubernetes 中 Alpine(musl) DNS 解析异常排查复盘 问题现象 某个 Kubernetes Pod 中: 报错: 但: 能够正常解析:…