Services

observability / running

Loki + Alloy

Central Docker log collection for all six Docker LXCs, surfaced through Grafana Explore instead of one-off SSH log hunts.

What it is

Loki is the log store and Grafana Alloy is the collector layer. Together they pull Docker logs from the homelab's Docker LXCs into one searchable place, with Grafana Explore as the UI.

Why I run it

Before this, service debugging meant remembering which LXC a container lived on, SSHing there, and tailing local logs by hand. That is fine once; it is not a system. Central logs make "what happened around the alert?" a Grafana question instead of a scavenger hunt.

How I use it

The first useful shape is operational search: filter by container, LXC, or time window when Uptime Kuma, Healthchecks, or a dashboard shows something odd. It is intentionally a private operator surface, not a public log viewer.