Blog

A half-remembered Mbps-to-bytes formula left the 3proxy bandlim cap at 5 Mbps instead of 40. The manual said bits all along.

TL;DR: A 3proxy bandwidth cap configured at 40 Mbps actually limited traffic to 5 Mbps because bandlim expects bits per second, not bytes, so the fix was changing 5000000 to 40000000. The same commit trimmed a GoAccess dashboard to three panels, surfacing the URL table immediately. Lesson: verify unit conversions against the manpage.

I tore out my self-built GoAccess stack for 3proxy's built-in admin counters, then realized counters can't answer which URLs eat the bandwidth.

TL;DR: The author deleted a self-built GoAccess dashboard for 3proxy, replacing it with the built-in admin directive and secure-mode counters. That answered how much traffic, but not which URLs or hosts consumed it. GoAccess returned properly: stripped fractional timestamps, filtered ADMIN lines, and nginx proxying WebSocket with explicit upgrade headers and longer read timeouts.

An audit of a news site with 9,200 articles in its sitemap and almost nothing indexed: crawl demand, internal linking, and the fixes that actually matter.

TL;DR: An audit of a news site with 9,200 articles in its sitemap found only about 44 that ever earned a Google impression, dropping to zero for the most recent publish months. Every technical check came back clean, so the problem is crawl demand: publishing volume far outpacing authority signals, plus an archive cut off from internal links. The fixes are editorial: lower the volume, build contextual links into old content, trim the news sitemap, and earn external validation.

A time-based bandwidth cap on a home LAN using Squid delay pools: bytes vs bits, the acl time trap, container TZ, and a second unlimited pool.

One folder or five scattered paths? The shape of your include set decides when subtree split is safe and when an rsync snapshot wins.

One WordPress install, two Git repos: a full Forgejo repo for recovery, a GitHub view repo carrying only source I wrote myself.

My per-commit dual-push ritual turned out more honest than a pull mirror that would quietly drop a 631 MB branch GitHub never had.

A local domain via /etc/hosts plus DOMAIN and ROOT_URL in Forgejo, and the 100 MB Cloudflare upload cap that makes the SSH path the sensible choice.

One Compose service driven by FORGEJO__ env vars: built-in loopback SSH, push-to-create, LFS, and a healthcheck that tells the truth.