Outage glossary

Plain-English definitions of the terms you run into during an outage — what “down,” “degraded,” a 503 error and “server status” actually mean.

Outage

A period when a service is unavailable or not working as intended for its users. An outage can be total (nothing works) or partial (some features fail). It ends when the service is restored.

Downtime

The amount of time a service is unavailable during an outage. Measured from when the problem starts to when it's resolved — the honest figure is how long users actually experienced it, which can be longer than the vendor reports.

Degraded service

The service is up but not fully working — slow, erroring intermittently, or with some features broken. It's the middle ground between 'operational' and 'down,' and it's often how an outage begins.

Server error (500, 502, 503, 504)

HTTP status codes a server returns when something's wrong on its end. 500 = generic internal error; 502 = a bad response from an upstream server; 503 = the service is unavailable or overloaded; 504 = an upstream timeout. Seeing these usually means the service, not you, is the problem.

Server status

Whether a service's servers are online and responding. 'Checking server status' means testing if the service is reachable and healthy — which is exactly what an outage tracker does.

Status page

An official page a company publishes to report its own incidents and uptime (often on Statuspage, Better Stack or Instatus). Authoritative, but it can lag the first minutes of an outage — which is why community reports matter.

Uptime

The percentage of time a service is available, usually over a month or year (e.g. 99.9% uptime ≈ 8.8 hours of downtime per year). The inverse of downtime.

Latency / brown-out

Latency is how long the service takes to respond. A 'brown-out' is when a service is slow enough to feel broken without being fully down — technically up, practically unusable.

DNS

The system that turns a domain name (instagram.com) into the server address browsers connect to. A DNS problem can make a perfectly healthy service unreachable — and a DNS issue on your network can make a site look down only for you.

CDN

A Content Delivery Network (like Cloudflare, Fastly or Akamai) sits in front of many sites to serve content faster. Because so many services share a CDN, one CDN outage can take down lots of seemingly-unrelated apps at once.

Maintenance window

Planned downtime a service schedules to make changes. Common for games and platforms (Fortnite, PSN), it's not an unexpected outage — but it still means the service is unavailable during that window.

Cascade / dependency outage

When one outage triggers others because services share infrastructure or an owner. If AWS or Cloudflare has a bad day, every app built on it can go down together. Mapping this 'what goes down with it' is DownRN's core idea.

Down for everyone or just me

The key question during an outage. If many people are reporting the same problem (and independent checks fail), it's the service. If only you're affected, it's usually your connection, DNS, device or network — try another network to confirm.

Downdetector

A well-known consumer outage tracker (a trademark of Ookla) that aggregates user reports. DownRN is an independent alternative that adds the cascade view and never fabricates report numbers.

Wondering about a specific app right now? See what's down right now or browse every app we track.