https://blog.margrop.net/en/tag/local-lvm/ Recent content in Local-Lvm on Margrop Blog Hugo -- gohugo.io en-US Tue, 05 May 2026 17:20:00 +0800 https://blog.margrop.net/en/post/proxmoxve-lvm-link-speed-uptime-kuma/ Tue, 05 May 2026 17:20:00 +0800 https://blog.margrop.net/en/post/proxmoxve-lvm-link-speed-uptime-kuma/ <blockquote> <p><strong>Short version</strong></p> <p>Proxmox VE <code>local-lvm</code> is usually an LVM-thin pool, not a normal mounted directory, so <code>df -h</code> is not enough to catch it before it becomes full. Network link speed is another quiet failure mode: services may stay online while a NAS or router uplink falls back to 100 Mb/s and turns large file transfers into a crawl. A practical fix is to feed both checks into the Uptime Kuma you already use: expose a small HTTP health endpoint for <code>local-lvm</code>, and push link-speed status from the Proxmox host with an <code>ethtool</code> script.</p> </blockquote> <p>This article walks through the design, the problems in a naive implementation, improved scripts, systemd units, Uptime Kuma configuration, verification steps, and safety notes. All addresses, tokens, host names, and environment-specific values are placeholders. Replace <code>&lt;PVE_HOST&gt;</code>, <code>&lt;KUMA_URL&gt;</code>, <code>&lt;PUSH_TOKEN&gt;</code>, and <code>&lt;INTERFACE&gt;</code> with values from your own environment.</p>