Help / Monitors
Ping monitor
A ping monitor checks that a server, router or other device answers ping from the internet. Use it when there is no website to load, or when you want to know the machine itself is reachable.
What you need first
- The host name or public IP address of the device, for example
server.example.comor203.0.113.5. - The device, and any firewall in front of it, must answer ping (ICMP echo) from the internet. Many servers and cloud firewalls block ping by default. If yours does, use a port monitor instead, or allow our checks through as described in Allowing our checks through a firewall.
Only public internet addresses work. Private addresses such as 192.168.1.10, and names like localhost, are refused when you save.
Set it up
- Open Monitors and press Add monitor.
- Under What should we watch?, choose Ping.
- In Host, type the host name or IP address. Do not add
https://or a path. - Give it a Friendly name you will recognise in an alert, such as "Office router".
- Choose how often to check in Check every. The list only shows the intervals your plan allows.
- Check Where alerts go at the bottom. It tells you how many alert channels will hear about an outage. If it says no channel is set, add one first (see Setting up alerts).
- Press Start monitoring.
The monitor page opens and shows "Waiting for the first check" until the first results come in.
How the check works
Each check sends three pings. If any one of them comes back, the check passes, so one lost packet on a long route is not an outage. The time shown is the fastest of the replies. If the host name has both an IPv4 and an IPv6 address, the IPv4 address is pinged.
On the monitor page, By location lists each checking city with its latest ping time, and the top of the page shows the lowest and highest ping and where they came from.
More settings
Open More settings on the form to change these. The defaults suit most people.
| Setting | Default | What it does |
|---|---|---|
| Down when this many locations agree | 2 | How many checking cities must fail before the monitor counts as down. |
| Each of them after this many failed checks in a row | 2 | Ignores one-off blips. Choose 1 to be told on the first failure, 3 for long outages only. |
| Check from | None ticked: all cities, taking turns | Tick cities to check from only those. The label says how many you can tick on your plan. |
| Alert when slower than (ms, optional) | Empty (off) | Sends a slow alert when the average ping of all cities goes over this. Tick Alert when any single location is slower to compare each city on its own instead. Slow alerts go to channels that have Slow ticked. |
| Timeout (seconds) | 10 | The longest a check may take, from 1 to 30. Each ping waits at most 3 seconds for its reply. |
| Repeat the down alert every (minutes, 0 = once) | 0 | Repeats the down alert until someone acknowledges the incident or the monitor recovers. Anything from 1 to 4 is saved as 5, the shortest repeat. |
| Send a "back online" message when it recovers | Ticked | Sends a recovery message to channels that have Back up ticked. |
For more on how cities vote, see How checks work.
Send a test
A monitor has no test button of its own. To check that alerts reach you, open Alerts and press Send test next to each channel. Tests send a real message.
To see that the check itself works, open the monitor and look at By location. Each city should show a ping time after the first round.
When it does not work
- "No reply to 3 pings" from every city. The device is off, or ping is blocked. Try pinging it yourself from a network outside your own. If ping is blocked on purpose, switch to a port monitor on a port the server uses.
- Only one or two cities fail. That is usually a routing problem between those cities and your network, not an outage. With the default settings it takes two cities to call the monitor down.
- "That address can't be monitored" when saving. The address is private or internal. Use the device's public address.
- It says down but the server works. See A monitor says down but my site works.
Still stuck? Email [email protected] and say which guide you were following.