What can PingFleet monitor?

HTTP and HTTPS endpoints, TCP ports, ICMP ping, TLS certificate expiry, domain registration expiry, heartbeats from cron jobs, metrics and thresholds from agents, and SNMP devices.

Do I need to open a firewall port for the agent?

No. The agent connects outward to PingFleet and reports its results. Nothing needs to be reachable from the internet.

Which addresses do the checks come from?

Every probe address is published at pingfleet.net/ips.json and pingfleet.net/ips.txt, so you can allow-list them in a firewall.

Is the free plan a trial?

No. It does not expire and does not ask for a card. It checks from one location; paid plans check from several and require agreement between them.

How long is history kept?

Individual check results are kept for 90 days, which is where the 90-day figures on status pages come from. Agent metrics are kept for 30 days.

Network intelligence

Uptime monitoring and network intelligence

Know what broke, where, and what it takes with it. Checks from independent networks, agents inside your own, and a live map of how it all connects.

Start free How it works

20 monitors free, for as long as you like. No card.

operational degraded down, and what it takes with it. Drag a node.

The map

The network map

A list tells you what is red. It does not tell you what that red thing sits behind, or which four services are about to follow it.

You place the nodes once. Only you know which of two boxes everything else leans on.

Live status on every node and link
Waypoints for the hops between two monitored things
An embed mode with no navigation, for a wall display

Open the live demo →

A PingFleet network map of a cloud estate: a CDN edge and WAF in front of an ingress per region, three services, a managed postgres shown in red with its read replica, redis, kafka with two workers, object storage and search.

Dashboard

Everything, on one screen

Availability across the last 24 hours, incidents per day, recent activity, the worst monitors, and the map. Hover an hour to see which monitors were down or degraded in it.

The PingFleet dashboard: monitor counts, system health, a 24-hour availability chart with a dip, and incidents per day.

Alerting

Alert escalation

Six ways to be told: email, SMS, Slack, Discord, Microsoft Teams, and a webhook to anything else you already run. Each destination is verified once and then reused.

Three consecutive failures: email to the ops mailbox and a message in #alerts. Still down ten minutes later: a text message to the on-call phone, but only between 08:00 and 20:00, Monday to Friday, in your timezone.

That is the entire configuration. It is one group, and it is attached to all forty monitors.

About alerting →

The PingFleet alerts screen: verified email, SMS, Slack, Discord, Teams and webhook destinations, above an escalation group with two steps.

The escalation

Steps, hours, and who is woken

A step names the destinations it fires, how long after the previous one, and the hours in which it is allowed to wake somebody. Nights and weekends are a checkbox, not a separate product.

An escalation group in PingFleet with two steps: email and Slack after three failures, SMS after ten minutes on weekdays between 08:00 and 20:00.

Incidents

Incident management

An incident opens the second a check fails, and stays private until you publish it. The words are drafted for you; you change them before anyone reads them.

Publish to one status page, several, or all of them
Subscribers confirm by email before anything is sent to them
Recovery is posted as an observation. Only you write "resolved"

A published incident thread in PingFleet: investigating, identified, monitoring and resolved, each with a timestamp and text.

Status pages

Public status pages

Every service and every incident is opt-in.

Your own domain, verified by DNS before it is served
Ninety-day uptime bars, from the stored checks
Two designs, chosen with a live preview

About status pages →

Agents

Agents for private networks

Databases on private subnets, an office NAS, a router behind CGNAT. The agent checks them locally and reports outward, with no inbound rules and no open ports, and sends CPU, memory, disk and network from the machine it runs on.

Linux, Windows, MikroTik RouterOS and Docker
One agent on the free plan, 25 on Pro

About agents →

Four agents in PingFleet and, for the selected one, CPU, RAM, disk and network graphs from the last hour.

Monitor types

Ten monitor types

Not one generic check bolted onto everything.

Website, ping and port checks
Certificate and domain expiry, on your thresholds
Cron heartbeats, SNMP, and agent metrics with thresholds

The PingFleet monitor list grouped by category, with a certificate monitor selected showing 43 days until expiry.

The fleet

The probe fleet

Checks run from separate providers on separate autonomous systems, so that two of our machines cannot share one network's bad day. Every address is published at /ips.json, as a file you can paste into a firewall rule.

Marked points are the machines that run checks. The arcs are checks leaving them for targets elsewhere, not places we check from.

Dependencies

What a dependency failure looks like

Watch it happen.

down degraded. This is the second the map earns its place.

Consensus

Consensus between locations

Questions

What can PingFleet monitor?: HTTP and HTTPS endpoints, TCP ports, ICMP ping, TLS certificate expiry, domain registration expiry, heartbeats from cron jobs, metrics and thresholds from agents, and SNMP devices.
Do I need to open a firewall port for the agent?: No. The agent connects outward to PingFleet and reports its results. Nothing needs to be reachable from the internet.
Which addresses do the checks come from?: Every probe address is published at pingfleet.net/ips.json and pingfleet.net/ips.txt, so you can allow-list them in a firewall.
Is the free plan a trial?: No. It does not expire and does not ask for a card. It checks from one location; paid plans check from several and require agreement between them.
How long is history kept?: Individual check results are kept for 90 days, which is where the 90-day figures on status pages come from. Agent metrics are kept for 30 days.