PylonMon watches your website, API, or server from outside your network — checking as often as every 30 seconds and paging you the moment it stops responding, so you hear about downtime before your customers do. And uptime is just the front door: the same account watches cron jobs, server vitals, and your Prometheus alerts too.
Free forever for 20 monitors · No credit card · No agent to install
# PylonMon checks your site from outside your network: GET https://your-site.com 200 OK · 142ms GET https://your-site.com 200 OK · 138ms # the moment it stops responding, you're paged: your-site.com is DOWN — connection timed out · confirmed on recheck
Paste your website, API endpoint, or server address. No agent, no code, no DNS changes — just the address you want watched.
PylonMon polls it from off your network, as often as every 30 seconds, checking the status code, response time, and — if you want — that the page actually contains the text it should.
The instant it can't reach your site, you're alerted — email, Slack, Discord, SMS, voice, webhook, PagerDuty or Opsgenie, with escalation ladders.
A monitor is only useful if you trust it. Before PylonMon pages you, it immediately re-checks a failed poll from scratch, and you can require a few consecutive failures — so a one-second network blip doesn't wake you at 3am, but a real outage still gets to you in seconds.
And a 200 OK isn't the whole story: a broken deploy can serve a blank white page that still returns 200. PylonMon can assert the response actually contains the text it should (or that an error string is absent), so you catch the outages a plain status check sleeps through.
On every paid plan, add multi-region confirmation: before a page goes out, the outage is verified from a second region. If our west-coast probe can't reach you but the east-coast one still can, that's a routing blip — you get an event in the timeline, not a 3am page. When both regions agree it's down, the alert says so: confirmed from 2 regions.
When something breaks you can publish a branded status page so customers self-serve instead of flooding your inbox.
Nobody wants three monitoring bills and four alerting inboxes. Uptime checks are the front door to PylonMon, but the same account — and the same alert channels, escalation ladders, and status page — covers the rest of your stack:
A dead-man's-switch for the things that should phone home: cron jobs, backups, batch pipelines. One curl at the end of the job — if it ever goes quiet, you're paged. Silence becomes a signal.
Cron & backup monitoring →One tiny open-source agent pushes CPU, memory, disk, and temperature — plus local HTTP/TCP/ping probes for the services an outside checker can't reach. Alert thresholds included on every plan, even free.
Server monitoring →Already running Prometheus? Point Alertmanager at PylonMon and your existing alert rules gain SMS, voice calls, and escalation — without you babysitting an inbox or running another pager service.
How alert ingest works →Email, Slack, Discord, Teams, SMS, voice calls, webhooks, PagerDuty/Opsgenie — with on-call ladders that keep escalating until a human acks. It can even page your AI agent with full incident context.
Every monitor & channel type →No — uptime is one monitor type. The same account runs cron/backup heartbeats, server vitals via the beacon agent, and Prometheus alert ingest, all feeding one set of alert channels and one status page. Most customers start with an uptime check and add the rest as they go.
No. Uptime checks run entirely from our side — you just give us a URL or address. There's no agent on your servers and nothing to maintain. (Server vitals are the one exception — that's what the optional beacon agent is for.)
Checks run as often as every 30 seconds (1-minute on the free plan). Alerts fire on the first confirmed state change, so you hear about a real outage in seconds — without repeat nagging once you've been told.
That's exactly what the confirmation re-check and consecutive-failure threshold are for. A blip won't page you; a genuine outage will.
Free forever for 20 monitors with 1-minute checks and HTTP/S, TCP & ping monitors. Paid tiers start at $9/mo for 30-second intervals, more monitors, SMS/voice, DNS, and certificate/domain-expiry monitoring. See pricing →
Set up in about two minutes. No credit card.