[Feature]: Auto cleanup of stale repeaters and tiles

[MrAlders0n]

Platform

MeshMapper Web

What feature do you want?

Add a per-region setting, configurable by regional admins, that automatically
cleans up repeaters offline for X days, along with the coverage tiles/cells
generated from those repeaters' data.

The threshold is selectable at 30, 60, or 90 days, defaulting to 60. The setting
is opt-in per region and defaults to OFF, so no region loses data unless an admin
deliberately enables it.

When enabled, a scheduled server-side job evaluates each repeater's last-seen
timestamp against the region's threshold. Repeaters past the threshold are flagged
stale and removed (or archived), and the coverage data they contributed is cleaned
up accordingly.

Why do you need it?

As regions mature, maps accumulate repeaters that were decommissioned, relocated,
or stood up as one-off test nodes. These stale repeaters and their leftover tiles
clutter the map, misrepresent current real-world coverage, and add noise to the
data. Cleanup is manual today, which does not scale as regions grow and as we
onboard more regional admins. A safe, opt-in auto-cleanup keeps regional maps
accurate without constant manual pruning.

How important is this?

High - Would help a lot

Anything else?

No response

[MrAlders0n]

Expanding the scope here a bit.

Alongside the repeater-tied cleanup above, I'd like this to also cover a general per-region ping-data retention period that a regional admin can set independently. So beyond removing tiles from dead repeaters, an admin could set something like a 90-day retention on all ping data in their region, ageing out anything older regardless of whether the source repeater is still alive.

This came out of a community request, and the reasoning is solid:

• Map trust: older regions are carrying pings from repeaters that have been dead a long time, and that volume only grows as an area matures. A retention window keeps the map reflecting current reality so people can trust what they are seeing.
• System health: clearing out old stale ping data is also better for the backend at large, keeping the ping tables from growing unbounded.

One thing to settle in design: a raw-ping retention purge needs to recompute or drop the coverage cells derived from those pings, otherwise we leave behind cells pointing at deleted data. Probably the same archive-first approach as the repeater cleanup.

Same shape as the repeater setting: 30 / 60 / 90 day windows, default 60, opt-in per region and off by default.

[Bjorkan]

Would like a 14 days window aswell for regions who want only the freshest data! 👀 but default 60 sounds good!

[MrAlders0n]

Would like a 14 days window aswell for regions who want only the freshest data! 👀 but default 60 sounds good!

It's just a variable at the end of the day I'm good with just putting a text box for regions to put whatever number they want in

[PappaNiklas]

Thanks for taking this to production, it will make the map closer to reallity (and clear up your database a bit)