The site went down, so a message lands in Telegram

You usually learn your site is down from a customer. Paying for monitoring over one landing page feels excessive, and checking by hand every morning does not happen. This scenario checks by itself and stays quiet while all is well.

What is inside

1. Scheduled trigger — the entry. There is no incoming message here: the scenario starts first, on an alarm clock.
2. Request to the site — an ordinary HTTP request with a ten-second timeout.
3. Message in Telegram hangs off the request’s Error output. Anything that did not work — a 500, a dropped connection, a domain that does not resolve — arrives here.
4. Warned — the exit.

Nothing is wired to the request’s success output, and that is not an omission. While the site answers there is nothing to do: silence is the good news. To confirm the scenario is alive, look at Runs — every firing is there, including the quiet ones.

What to set up for yourself

• The address in the Request to the site node: replace REPLACE_ME with your own. Prefer a page that touches the database over the home page: home is often served from cache and stays green when everything behind it has stopped.
• The chat in the send node: replace REPLACE_ME with the id of the chat or channel to write to. You can learn yours from any bot such as @userinfobot.
• The message text. It currently carries the time of the check: {{ nodes.entry.data.cron.fired_at }}.
• The interval is not in the graph: it is in the schedule settings on the Publish tab.

What to connect

You need a Telegram bot. Create one with @BotFather, take the token, add a connection in settings, then pick the bot in the publish targets. After that enable the Schedule target and set an interval.

One limit is worth knowing up front: on the free plan a schedule fires at most every thirty minutes. Set five and the scenario keeps working, just less often than it says. That is a plan ceiling, not a fault.