https://blog.margrop.net/en/tag/intel/ Recent content in Intel on Margrop Blog Hugo -- gohugo.io en-US Sat, 13 Jun 2026 08:30:00 +0800 https://blog.margrop.net/en/post/photoprism-init-intel-stuck/ Sat, 13 Jun 2026 08:30:00 +0800 https://blog.margrop.net/en/post/photoprism-init-intel-stuck/ <blockquote> <p><strong>TL;DR (The Short Version)</strong></p> <p>The <code>PHOTOPRISM_INIT: &quot;intel&quot;</code> line in PhotoPrism&rsquo;s Plus Docker image is <strong>not</strong> just a flag to enable Intel hardware acceleration at runtime — it&rsquo;s a one-shot <strong>system-level installer</strong> that runs <code>apt-get dist-upgrade</code> against <code>archive.ubuntu.com</code> and then installs 8 Intel/VA-API packages before the main process is allowed to start. On a fresh container this takes <strong>5–10 minutes</strong> (sometimes more), during which nothing inside the container listens on port <code>2342</code>.</p> <p>Meanwhile: <code>docker ps</code> says <code>Up</code>, Portainer is green, <code>ss -ltn</code> on the host shows <code>:2342</code> LISTEN — but your browser gets <code>ERR_CONNECTION_RESET</code> / <code>Connection reset by peer</code>. Because host-side <code>docker-proxy</code> accepts the connection and then tries to forward to a port inside the container that has no listener, the kernel sends back an RST.</p> <p><strong>Fix:</strong> just remove or empty that one line. If you already pass <code>/dev/dri/renderD128</code> via <code>devices:</code>, the host&rsquo;s drivers and userland libraries are perfectly usable from the container — there&rsquo;s nothing to &ldquo;install&rdquo; in there.</p> </blockquote> https://blog.margrop.net/en/post/n5105-in-promox-ve-frequently-dead/ Sat, 04 May 2024 07:55:22 +0800 https://blog.margrop.net/en/post/n5105-in-promox-ve-frequently-dead/ 1. Problem background Recently, major PT websites were developed and registered on May Day, so I registered several websites, and then downloaded and uploaded them like crazy in order to complete novice tasks. As soft routers, N5105 and iKuai, when under high load for a long time, the virtual machine crashed randomly, and the CPU remained at 100% for a long time. The virtual machine could only be killed and restarted, which greatly affected the completion of novice tasks. https://blog.margrop.net/en/post/proxmox-ve-graphics-card-enable-intel-gvt-g/ Sat, 23 Jan 2021 11:23:04 +0800 https://blog.margrop.net/en/post/proxmox-ve-graphics-card-enable-intel-gvt-g/ <blockquote> <p>Note: Intel GVT-g is currently only supported by fifth, sixth, and seventh generation Intel Core processors and Xeon E3 v4/v5/v6 processors on PVE. My tests show that 8th generation and newer CPUs are not supported for now -_-! If this changes, feel free to contact me. Reference: <a href="https://pve.proxmox.com/wiki/PCI(e)_Passthrough">https://pve.proxmox.com/wiki/PCI(e)_Passthrough</a></p> </blockquote>