Every machine reports its own backups

The Phlo Dashboard gained a fleet-wide backup view: one card per node showing when its last run happened and exactly which files moved.

The reason it exists is the least glamorous one in this log. A nightly backup that quietly stopped running looks exactly like a nightly backup that runs fine, right up until the day you need it. Checking by hand means logging into each machine, and a check you have to remember to do is a check that stops happening.

So the node reports it. fleet::collect() already sends each machine's errors and metrics up to the dashboard, and the backup log rides along in the same feed: the most recent run, the number of files copied, the number deleted, and the paths themselves. Every node carries it, including headless ones with no dashboard of their own, because those are precisely the machines nobody looks at.

The dashboard renders one collapsible card per node, summarised as when it ran and how much moved, with the full path list inside. Two small pieces of care: malformed lines in the log are skipped rather than shown as entries, and a very long list is capped with a count of the remainder, because a card that takes ten seconds to open is a card nobody opens twice.

It sits behind the same admin gate as every other module, checked where the modules are resolved rather than only hidden in the menu.

Phlo Dashboard 1.0.1. See the Dashboard handbook.