Staleness

Scanopy marks an entity Stale when discovery hasn't observed it within its network's staleness window. Staleness is a signal that something may have gone quiet — a device powered off, a container removed, a host that moved — surfaced without deleting anything, so you decide what to do about it.

Stale does not mean removed. A stale host might simply be switched off, on the road, or briefly unreachable during a scan. Scanopy never asserts that an entity is gone; it only reports how long it's been since discovery last saw it.

The staleness window

Each network has its own window — how long an entity can go unobserved before it's flagged. Staleness is only meaningful relative to how often a network is scanned: a segment swept every 15 minutes and one swept monthly need very different thresholds, so the window lives on the network, not the whole organization.

The default window is 28 days. Set a network's own window in the network edit modal (Assets > Networks → select the network → Consider entities stale after), entered as days and hours. Set it to comfortably exceed the network's scan interval so a normal gap between scans never trips it. Leave it blank to use the default.

See Networks for where this fits among a network's settings.

What can go stale

Only entities that discovery manages can go stale — hosts, services, and subnets that discovery found and keeps re-observing. Anything you created by hand never goes stale, because Scanopy has no scan timestamp to judge it against and won't flag your own curated records.

Every entity is judged on its own last-seen time, independently of its parent. A service on a host that's gone quiet is only stale if that service itself hasn't been seen within the window — a recently observed service on an otherwise-dark host reads as current, and vice versa. This keeps a badge from ever contradicting its own "last seen" detail.

Where it shows up

Lists. Hosts, services, and subnets each carry a Stale badge on rows past their window, and a Stale only filter to narrow the list to them. You can also sort by last-seen to bring the longest-unobserved entities to the top.

Topology. Stale nodes carry a Stale tag on their card, with a "last seen" detail naming the entity and how long ago. A By staleness filter (Current / Stale) is available on the perspectives that show these entities, letting you fade everything except what's fallen behind. See Customizing Views.

The tag is amber, not red — matching the daemon status convention where red means broken and amber means behind. A stale entity may be perfectly healthy and simply unobserved, so the tag reads as "worth a look", not "error".