Five-stage timing
DNS, TCP handshake, TLS negotiation, time to first byte, transfer. Five numbers instead of one, so “it got slow” has a cause rather than a feeling.
Uptime monitoring with the diagnosis attached
Your site is down. Here is why.
Every other monitor hands you a status code and leaves the diagnosis to you at three in the morning. This one records the full picture on every check — five-stage timing, the certificate, the address it connected to, where the redirects went — and puts it in the alert.
No card. Five monitors, permanently — plus 14 days of the paid plan to try.
https://example.com/
Total
247 ms
- DNS
- 12 ms
- TCP
- 24 ms
- TLS
- 58 ms
- TTFB
- 121 ms
- Transfer
- 32 ms
Certificate valid for 62 days
Let's Encrypt · fingerprint unchanged
The problem
An alert that says “down” starts the work. It does not do any of it.
You get the message at 03:14. Then you open a terminal, because the message told you nothing you can act on. You dig the record. You curl with a write-out format you have to look up. You check the certificate. By the time you know whether it is DNS, the certificate, the origin or the CDN, it is 03:40 and you have learned five things your monitor already knew and did not say.
All of it was in the check. It should have been in the alert.
How it works
Three steps, and the first one takes a minute.
-
01
Add a URL
Paste it and press preview. We fetch it once, immediately, and show you the status, the timing and the certificate before you save anything. If we cannot reach it, you find out now rather than in an alert tomorrow.
-
02
We check on a schedule
Every fifteen minutes on the free plan, every minute on Pro, once you have verified you control the domain — until then Free and the free trial check every thirty minutes, and Pro every five. Each check records the whole picture, not just whether the status code was 200, so an alert can compare it with the last check that worked.
-
03
We tell you why
Email, Slack or a webhook, carrying the breakdown and the difference. Not “example.com is down”, but which stage got slow or where it stopped, which certificate or address changed, and where the redirects ended.
What we record
Six things on every website check.
This is the part most monitors charge extra for or do not do at all. It is on every plan, including the free one — because depth is the product, not the upsell.
TLS certificate
Issuer, expiry and fingerprint. An expiry you can plan around, and a fingerprint that changed when nobody deployed anything. On Pro, that change is a message on its own.
DNS
The address every check connected to and how long the lookup took, plus the A, AAAA, CNAME, MX and TXT records, refreshed every few hours — so a propagation problem looks like a propagation problem instead of an outage.
The redirect chain
How many redirects there were, where they ended and how long they took. Each hop is re-validated before we follow it, so a monitor can never be pointed at somewhere it should not go.
Your definition of healthy
Status code, a keyword that must or must not be in the body, a response-time ceiling, days left on the certificate. A page that returns 200 while rendering an error is still broken, and you get to say so.
What changed
When something breaks, the alert compares the failing check with the last one that worked and shows you the difference, rather than making you find it.
The part that matters at 3am
We do not wake you for one bad packet.
A single failed probe never sends anything, ever. A failure has to reach the threshold you set, and then a second, independent confirmation probe has to fail as well before an incident opens. Recovery is confirmed the same way. It is the whole difference between a monitor you trust and a monitor you mute.
Planned work is not an incident
Schedule the window and the deploy stops paging you. The checks keep running and keep recording — you just do not get shouted at for something you did on purpose.
Credentials stay encrypted
Anything you give a monitor to authenticate with is encrypted at rest and never comes back to the browser, never appears in a log, and never lands in an error message.
Plans
Five sites, monitored properly.
Free is not a trial. It does not expire, it does not ask for a card, and it is not a crippled version of the product — it runs the same probe, records the same six things and sends the same alerts.
Free
€0 forever
- 5 monitors
- Checks every 15 minutes
- Full depth on every check
No card. It does not expire.
Pro
€19 per month
or €190 a year — 2 months free
- 50 monitors
- Checks every minute
- Slack, webhooks and maintenance windows
Full Pro for 14 days, no card. After that, subscribe or stay on Free — nothing is deleted either way. Prices exclude VAT.
Questions
The things people ask first.
Is the free plan actually free?
Yes, and permanently. Five monitors, checks every fifteen minutes on a domain you have verified (thirty until you do), ninety days of history and email alerts, with no card and no expiry. What Pro sells is frequency and alerting — one-minute checks, Slack, webhooks, maintenance windows — not the depth.
Will it wake me up for nothing?
It is built specifically not to. One failed probe never notifies anybody. The failure has to reach your threshold and then be confirmed by a second, independent probe before an incident opens at all.
Can I monitor an API rather than a website?
On Pro, yes — custom methods, request headers, a request body and Basic or Bearer authentication, with assertions against the response. On Free you get GET requests and a status-code or keyword assertion.
Can you monitor something that is not a website at all?
Yes — a database, a mail server, an SSH daemon, a game server. A port monitor opens a TCP connection to any host and port and records whether it answered and how quickly, on every plan including Free. It is table stakes rather than the depth above, and worth saying so plainly: a socket returns no page, so there is no timing breakdown, no certificate and nothing to assert against. What it does answer better than an HTTP check is refused versus timed out — the service is dead, or the firewall is.
How quickly will I know?
On Pro, with the domain verified, a check runs every minute and the confirmation probe runs immediately after a failure, so a genuine outage is normally in your inbox inside two minutes. Before you verify the domain, Pro checks every five minutes, or every thirty during the free trial. On Free the interval is fifteen minutes once verified and thirty before.
How long do you keep the data?
Ninety days of full check history on both plans, then it ages out. Every check belongs to exactly one monitor and one account — nothing is ever shared between accounts, which is also why deleting a monitor genuinely deletes its history.
Add your first site in under a minute.
Paste a URL, watch us fetch it, and see the whole breakdown before you decide whether to keep it.
Start free with five sites