Synology Subdomains: SSL, Reverse Proxy, and Dynamic DNS
A subdomain that wouldn't resolve. An SSL certificate that existed but didn't work. Three hours of debugging Synology's certificate management to learn it was never properly assigned.
Browse 12 articles in this category
A subdomain that wouldn't resolve. An SSL certificate that existed but didn't work. Three hours of debugging Synology's certificate management to learn it was never properly assigned.
Four days, 463 messages, and one reverse proxy that refused to cooperate. The story of how I containerized my entire infrastructure by sheer stubbornness.
I self-hosted RustDesk to replace TeamViewer. Then spent three days debugging why I couldn't connect to anything from inside my own network. NAT hairpinning strikes again.
A week of making the Gentoo build swarm actually reliable. Runaway drones, dead cron jobs, self-healing systems, and the satisfaction of watching it all come together.
Storage archaeology, dashboard building, a power crisis, log floods, network ghosts, phone-guided Proxmox recovery, and the documentation sprint that made it all worth it. The complete story.
Three months ago, a rogue drone failed the same build 500 times in a minute. Today, the swarm self-heals in 60 seconds. This is v1.0.
I restarted a container and every free-tier LLM provider went down at the same time. Groq, Cerebras, SambaNova, OpenRouter, NVIDIA NIM — all exhausted in a chain reaction that started with a single config flag.
A 1,556-package @world update crashed while KDE was running. The emerge completed on its own — but libxcrypt lost SHA512 support, /etc/shadow got corrupted, and login was impossible. The fix took 14 days, a dual-boot chroot, and a 28KB automated recovery script.
Moving from manual SSH and shell scripts to declarative infrastructure management using Ansible playbooks and Terraform.
The complete timeline of my homelab journey. ESXi to Proxmox, single drives to redundant arrays, VPNs to mesh networks. What worked, what broke, and why I'm still doing this.
Running Proxmox VE at a remote location. LXC containers, ZFS storage, VPN connectivity, and managing infrastructure you can't physically touch.
Deploying Prometheus, Grafana, Loki, cAdvisor, Smokeping, Healthchecks, and Promtail on one box. Seven containers, 45 minutes, and one ironic nodejs detour.