Ping (ICMP) Monitoring
A Ping monitor sends ICMP echo requests to a host and reports what came back: how many packets were lost, how long the round trip took, and how much that time varied. It is the lowest-level check PingZen offers — it asks whether the machine is there, not whether anything useful is running on it.
What It Is For
Servers, routers and network gear
Anything with an address that answers ICMP: a bare VPS, a router, a NAS, a printer. No agent, no open service port, nothing to configure on the far end.
Link quality, not just reachability
Packet loss and jitter say more about a shaky connection than a plain up/down. A link that answers every request but loses one packet in twenty is worth knowing about before users complain.
Separating "the network" from "the application"
Run a ping monitor next to an HTTP one on the same host. If the site is down but ping is fine, the box is up and the application is not.
What It Is Not For
ICMP answers one question. These need other monitors:
| You want to know | Use instead |
|---|---|
| Is the website or API actually working? A host answers ping perfectly while its web server is dead | HTTP / HTTPS monitor |
| Is a specific port open and accepting connections? | TCP monitor |
| The host ignores ICMP — many cloud providers, firewalls and CDNs drop it by policy, so a ping monitor would report a permanent outage on a perfectly healthy host | TCP monitor on a port it does answer, or an HTTP monitor |
| Does the certificate expire soon? | TLS/SSL certificate monitor |
| How fast does the page load for a visitor? Round-trip time is not page load time | PageSpeed monitor, or the response time on an HTTP monitor |
What It Measures
Every check sends several echo requests and records:
- Packet loss — the share of requests that got no answer
- Round-trip time — minimum, maximum and average across the packets that did answer
- Jitter — how much the round-trip time varied between packets, which is what makes calls and video stutter even when the average looks fine
Status Logic
By default a ping monitor is simply up or down: thresholds are off until you set them, and a value of zero means disabled.
| Result | Status |
|---|---|
| Packets came back, no threshold exceeded | UP |
| Loss or round-trip time above the warning threshold | WARNING |
| Loss or round-trip time above the critical threshold | DEGRADED |
| Every packet lost, or the host did not resolve | DOWN |
A degraded ping monitor is not an outage: the host is answering, just not well. It opens no incident and costs no uptime.
Configuration
Host
A hostname or an IP address. No port — ICMP has none.
Packet count
How many echo requests per check. Four by default, one to ten. More packets make the loss figure meaningful; one packet can only ever report 0% or 100%.
Packet size
64 bytes by default, from 32 to 1500. Larger packets can expose MTU and fragmentation problems that small ones hide.
Packet loss thresholds
Warning and critical percentages. Off by default. On a link you expect to be clean, a warning at a few percent catches trouble early.
Response time thresholds
Warning and critical round-trip time in milliseconds. Off by default. Set them from a measured baseline rather than a guess — see how we measure latency.
Key Features
- Packet loss, min/max/average round-trip time and jitter on every check
- Separate warning and critical thresholds for loss and for latency, both optional
- Degraded and warning states that do not count as downtime
- Works against anything that answers ICMP, with nothing to install on the target