uptime monitoring · opening soon
We poke your servers. You get to sleep.
You Up? checks your sites, APIs and cron jobs from locations around the world, and only wakes you when most of them agree something is actually down.
Questions before then: support@uup.dev
No card. Every feature on every plan. First result in under 60 seconds.
how it works
One grumpy location doesn't get to wake you up.
When a check fails, we retry, then ask other locations to look. An Incident opens only when the majority agree. Then you get one message, not twelve.
Retry. A blip isn't an outage.
Ask around. Other locations check the same URL, right now.
Agree, then tell you. One outage, one message. Follow-on alerts fold under it.
Free probes from one location, so it retries before it tells you. Paid plans probe from five and ask around.
Incident opened · 3 of 3 locations agree
Moments after the first failure. One message sent to Slack and email.
then, at 03:14
Where it stopped. Never a guess.
When your site goes down, we tell you where the request stopped (the name, the connection, the secure handshake or your server's answer) and give you a report to send to your hosting provider.
api.acme.com did not answer.
3 of 3 locations agree
Where it stopped
The name, the connection, the secure handshake or the answer: you see which step it reached and what came back.
A report for your hosting provider
Addresses, networks, timings and hops in one copy, ready to paste into a support ticket:
No answer came from 203.0.113.7 on port 443. A router at 198.51.100.1 (Example Transit, AS64500) was the last to answer, 4 hops out.
Down for some visitors
We try IPv4 and IPv6 on their own, so you see when one answers and the other doesn't.
last 24 hoursnow
Regional outage, Frankfurt, 18 min. Your Monitor stayed ok. Nobody was paged. It's on the record if you want to look.
regional outages
A bad night in Frankfurt isn't your outage.
When one location can't reach you but the rest can, we mark it in amber and let you sleep.
four checks per Monitor
Awake isn't the only question.
- uptime
Is it answering at all?
200 OK · 142 ms - performance
Is it answering fast enough?
slower: waiting for the answer - certificate
Is the padlock about to fall off?
expires in 12 daysSSL and domain monitoring → - domain
Did someone forget to renew?
renews in 41 daysSSL and domain monitoring →
- http
Methods, headers, auth, expected codes, and “does the body say ok?”
- tcp
Knock on host:port. Databases, mail, game servers.
- ping
Plain ICMP. We tell you up front if a host filters it.
- heartbeat
Reversed: your cron job texts us. Quiet too long? “No signal for 1h 11m.”
Cron job monitoring →
api.acme.com · 90 days
99.97%
Not 100.00%. Because it wasn't.
We never round downtime away. Hours we couldn't measure say “not measured”. If we get one wrong, it stays on the record, marked “reported in error”.
- ok
- Incident
- regional outage
- not measured
Channels
Tell us where to find you.
Route each Project's alerts to the right people. Every Channel is verified before we rely on it, and every delivery is in the log.
Every feature on every plan. Plans only differ in counts.
- Free
- Small
- Medium
- Large
- Extra Large
Go to bed. We'll stay up.
Questions before then: support@uup.dev