Fewer alerts, and the facts behind each one

One notice for a flapping monitor, and the facts behind it

When a monitor keeps switching between up and down, UptimeWarden opens one flapping incident and sends one notice that does not page anyone, instead of an alert for every drop. The monitor's card shows when the flapping started, how many times it went down and for how long, which errors came back (timeouts, DNS failures, 5xx responses, refused connections) and whether other regions saw the same. If the monitor stays down, you get a normal down alert.

On Professional and Enterprise, the card adds a suggested reason written from those facts.

Limits:

The suggested reason is a lead to check, not a diagnosis.

Plans:

Every plan. Suggested reason on Professional and Enterprise.

One message instead of forty when monitors fail together

When several monitors fail at the same time, their email, Slack, Teams and webhook alerts are collected into one message. It lists the monitors and what they have in common: the same error, the same host or IP address, the same region, or a dependency you set up. SMS and voice calls go out right away and are never held for grouping.

On Professional and Enterprise, the message can include a likely-cause hint written from the errors your monitors returned. If the hint is slow or unavailable, the message goes out without it.

Limits:

Monitors that fail together do not always fail for the same reason. The message shows what they share and leaves the conclusion to you.

Plans:

Every plan. Likely-cause hint on Professional and Enterprise.

Which services go down together, and why we think so

UptimeWarden looks back over the last 30 days for monitors that keep failing in the same minute, and for monitors that share infrastructure: the same IP address, CDN or TLS certificate. Each link it finds says why it is there, such as "went down together 12 times", "same IP" or "you added it". You can confirm a link, reject it, or turn it into a dependency.

Limits:

Failing together is a hint, not proof: two monitors can fail in the same minute for different reasons. That is why every link shows its evidence and waits for your answer.

Plans:

Every plan. The service map is on Starter and above.

A warning before a slowdown or a failed certificate renewal turns into an outage

Some problems build up for hours or days before anything goes down. UptimeWarden sends a notice when a monitor's response time stays well above its usual level, when a certificate that should have renewed by now has not, and when the share of failed checks has been rising for several days. On Professional and Enterprise, SLO alerts also tell you when an error budget is burning faster than it can last.

Limits:

Early warnings do not page anyone. Sudden outages, such as a crash or a bad deploy, give no warning and cannot be predicted.

Plans:

Every plan. SLO alerts on Professional and Enterprise.