- Online now
- 70.6%
- Reboots/day
- 1.3
- Fleet
- 50k+
Internet weather report
A daily read on the public networks the remote.it fleet connects over —
Tuesday, July 21, 2026, worldwide. Conditions: stormy.
Not a remote.it status page — a snapshot of the local networks our devices connect over. “Stormy” means rough internet in that region, not a remote.it outage.
Four forecasts, one fleet.
The whole fleet, then each way it connects. Connectivity leads; how a device is used colors the rest — see the notes below.
- Online now
- 57.0%
- Reboots/day
- 1.4
- Fleet
- 50k+
- Online now
- 74.9%
- Reboots/day
- 1.1
- Fleet
- 10k+
- Online now
- 81.7%
- Reboots/day
- 1.3
- Fleet
- 50k+
The last seven days.
Global average connectivity, one point per day. History builds forward — the report keeps a rolling daily series.
Average connectivity across the reporting fleet, one point per day.
| Date | Average connectivity | Conditions |
|---|---|---|
| Monday, July 20 | 92.0% | Partly cloudy |
| Tuesday, July 21 | 74.1% | Stormy |
Connectivity by country.
Each country shaded by its observed average connectivity. Grey means no reporting fleet today.
Connectivity by state.
All 50 states and DC shaded by connectivity — Alaska and Hawaii inset at lower left. Territories appear in the table.
How we grade the weather
We grade a connection by how often it drops while the device is online — reconnects per online-hour — not by raw connectivity. That way a device that sleeps to save power isn't called stormy just for being offline by design. Clear is rock-solid, Partly cloudy the occasional reconnect, Stormy frequent drops. The bar on each row is that grade mix; the percentage beside it is the region's average time connected — a separate axis.
Today, and the last 24 hours
Each report covers the last 24 hours of devices seen — today's active fleet, not a rolling week — so the trend actually moves day to day. Because we normalize per online-hour, a region can read as reliably connected yet still show lower connectivity when its devices duty-cycle. We report cellular vs. Starlink as an observed difference, not a verdict — usage and power still color it.
Confidence, privacy & sources
The Sample dots show how much data backs a row — a larger device sample means a more reliable average (we never publish exact counts). For the same reason we rank countries by a sample-adjusted connectivity: a small region's average is pulled toward the global mean by its sample size, so a lucky handful of always-on devices can't top a large, genuinely-excellent region. The number shown is always the region's real observed connectivity — only the ordering is adjusted. Aggregates only: any region with fewer than 20 devices is omitted, and a network-class row within a region needs at least 5 devices — below that, an “average” is close to one device's raw telemetry; no counts, IPs, or user data are ever published. World map: “Simple World Map” by Al MacDonald / Fritz Lekschas, CC BY-SA 3.0. US geometry: public-domain US Census cartographic boundaries (us-atlas).
Generated 2026-07-21T00:31:16.901Z · powered by remote.it