ICMP / Ping monitoring

Ping what cloud monitors can't — and see more than a single number.

Cloud-based ping tools reach the public internet. Full stop. CuliPulse lets you run an agent on your own network so you can ping any host that lives there — routers, switches, internal servers, database hosts — with no public exposure required. And instead of one RTT number, you get packet loss, min/avg/max, and jitter, so you know the real story.

Start free See the metrics
No credit card · free plan forever
gateway-switch-1 · ICMP
12 of 12 packets received · 0% packet loss
upagent "prod-network" · 60s interval
Private network host — reachable from your agent, invisible to the public internet.
More than a single number

Packet loss, RTT breakdown, and jitter — not just "up" or "down".

A single average RTT hides a lot. A host that responds in 18 ms on average might be dropping 20% of packets and spiking to 90 ms on bad runs. CuliPulse surfaces the full picture every check cycle so you can see degradation before users feel it.

gateway-switch-1 · healthy
packet loss0%
packets sent / received12 / 12
avg RTT18 ms
min / max RTT9 ms / 24 ms
jitter±4 ms
0% loss · 12/12 · 18 ms avg · 9–24 ms · ±4 ms jitter
db-host-3 · degraded
packet loss25%
packets sent / received12 / 9
avg RTT67 ms
min / max RTT14 ms / 203 ms
jitter±58 ms
High packet loss and jitter — degraded alert sent.
Alert before it's fully down

Flag high packet loss or latency as degraded — don't wait for a complete outage.

A host that's dropping 30% of packets or spiking to 200 ms is not "up" in any meaningful sense. CuliPulse can treat that as a degraded state and alert you immediately — so you're investigating while users are only slightly affected, not scrambling after the fact.

Packet loss threshold

Set a maximum acceptable packet loss percentage. Cross it and the monitor goes degraded — a yellow alert, not red, so you can triage before it's critical.

loss > 10%loss > 25%custom %

RTT threshold

Set a max average round-trip time. When latency climbs past your threshold the monitor flags degraded so you know something is wrong on the path.

avg > 50msavg > 100mscustom ms

Jitter threshold

Erratic round-trip times — high jitter — are a sign of an unstable path even when the average looks fine. Set a jitter limit and catch it early.

jitter > 20msjitter > 50mscustom ms
Reach what cloud can't

Ping private hosts on your own network — routers, switches, internal servers, anything.

Every cloud-based ping monitor pings from the public internet. That's fine for a public website — useless for the router in your rack, the database host that never has a public IP, or the internal server your team actually depends on. CuliPulse lets you run a small agent on any network. That agent pings from inside, so you watch what your users actually experience.

An agent on your network

Install CuliPulse's lightweight agent on any machine that can reach your hosts. It runs the ping checks from there, reports results back, and needs no inbound firewall rules — just outbound HTTPS.

Linuxamd64 · arm64Docker or systemd

No public exposure needed

Private hosts stay private. The agent initiates all traffic outbound. There's no need to open ports, poke holes in your firewall, or expose anything to the public internet.

internal IPsprivate subnetsair-gapped segments
CuliPulse
🟡 core-router-2 (ICMP) is degraded
Packet loss 28% · avg RTT 94 ms (threshold: 50 ms)
From agent "datacenter-1" · last checked 30s ago
Fewer false alarms

Confirmed from several checkpoints before you're paged.

Add more than one checkpoint — a mix of CuliPulse's network and agents on your own infrastructure — and we wait for several of them to agree before calling something down. A single ICMP timeout on one path won't wake you up at 3 am.

db-primary-1 · 2 of 3 agree
agent prod-1 100% loss
agent prod-2 83% loss
CuliPulse network 0% loss
2 of 3 checkpoints agree it's unreachable → alert sent. A transient ICMP drop never pages you.
CuliPulse
🔴 db-primary-1 is down
Ping unreachable · confirmed from 2 agents
Packet loss 100% on prod-1, 83% on prod-2

Start pinging in 2 minutes.

Add a ping monitor, point it at any host your agent can reach, set your loss and latency thresholds, and get alerts you can trust — even for hosts that never touch the public internet.

Start free