Skip to content
Tech News
← Back to articles

The First Ten Minutes of a Big Outage: What Actually Helps

get Anker 5G LTE Mobile Hotspot → more articles
In Short

When a service everyone uses stops working, most of the first ten minutes get spent on steps that could not possibly have helped. A short fixed routine tells you whether the problem is yours, and it takes about ninety seconds.

The reflex, when something stops working, is to fix it. During a widespread outage that reflex is expensive, because every repair you attempt lands on a system that was never broken. People reboot a router that is fine. They sign out of an app and discover they cannot sign back in. They reset a password at the exact moment the service that stores passwords is refusing writes. An hour later the provider posts that it was a configuration change in one region, and the only lasting damage on your side is self-inflicted.

A fixed routine avoids all of that, and it is short.

Start with a second device on a different network. Take a phone off Wi-Fi, onto cellular, and try the same thing. If it fails there too, nothing on your network is the cause and nothing you do to your network will help. This one step is the whole diagnosis most of the time, and it is the step people skip because it feels passive.

Next, find out whether the whole service is down or one function of it. Mail failing while photos sync normally is a useful clue: you are probably looking at an authentication or single-subsystem fault rather than a region falling over. It also tells you how long to expect. Partial failures inside one product tend to resolve faster than the kind where every request to a provider times out.

Then check the status pages, and hold them loosely. A status page is updated by a human being who works at the company, usually after an internal severity call, which is why it stays green for a while after your software has clearly stopped working. Crowd-sourced report maps move faster, but they are a count of people complaining, and they will show a spike on one service because a different service that it logs into is the actual problem. Both are evidence. Neither is an authority.

The one thing to avoid is the thing an outage most tempts you into: changing credentials. If sign-ins are failing, the identity layer is a prime suspect, and identity layers are exactly where a half-committed change does lasting harm. A password reset that writes to a primary but not to the replica, a two-factor re-enrollment that registers a new secret the server never stores, a recovery flow started and abandoned - these produce the accounts that still will not open tomorrow, after everyone else is back.

Last, write down the time and what you saw. If the outage cost you a transfer that posted twice, a flight you could not check into, or a filing deadline, a timestamped note is worth more than your memory in three weeks when you are asking someone to reverse a fee.

The hard part of this routine is that most of it rewards doing nothing. That is not a satisfying posture during an outage, but it is usually the correct one, and it is the only approach that cannot make your afternoon worse than the outage already has.

Key Takeaways
Worth a Look

Anker 5G LTE Mobile Hotspot — When an outage hits, having a cellular-based device ready to cross-check your home network makes diagnosis instant instead of guesswork. A dedicated hotspot gives you a clean, separate network path so you can quickly confirm whether an issue is on your end or the provider's. It's a handy tool to keep charged for exactly the ten-minute triage this article describes.

See Anker 5G LTE Mobile Hotspot on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.
Keep reading
downdetector 30 briefs → outages 19 briefs → troubleshooting 15 briefs → two-factor authentication 15 briefs →