New: Connect AI assistants via MCP

Core Concepts

Ping Monitoring

Ping Monitoring (ICMP Echo)

Ping monitoring sends ICMP echo requests to a host from each of your monitored regions and reports whether the host responds. It's the right check type for any server or network device that needs to be verified at the network layer — particularly hosts that don't speak HTTP, such as database servers, load balancers, switches, routers, IoT devices, or internal infrastructure.

Use Cases

  • Database servers — confirm a Postgres or MySQL host is reachable without exposing a port to the public
  • Network appliances — routers, firewalls, switches, and other devices that respond to ICMP but don't run a web server
  • Cache and queue servers — Redis, Memcached, RabbitMQ hosts where the protocol port is firewalled
  • IoT and edge devices — anything you just need to know is "alive on the network"
  • Reachability sanity check alongside HTTP — create a second monitor of type Ping for hosts that also serve HTTP, to distinguish "network down" from "application crashed"

How It Works

Every check interval, from every monitored region, the system:

  1. Extracts the hostname from the monitor's URL (a full URL like https://db.example.com works just as well as a bare hostname like db.example.com)
  2. Sends three ICMP echo packets with a 10-second overall deadline
  3. Records success if at least one packet returns, along with the average round-trip latency
  4. Persists a regional result so the multi-region consensus engine can decide whether to alert

Setting Up Ping Monitoring

  1. Click Add Monitor and select Ping as the monitor type
  2. Enter the hostname or full URL of the host you want to ping
  3. Pick the regions you want the host pinged from — consensus across regions makes the alert decision
  4. Save the monitor — the first check runs immediately and every subsequent check interval thereafter

Ping is a top-level monitor type, not a sub-check. A single monitor is either a Ping monitor or an HTTP monitor — create separate monitors if you need both kinds of check against the same host.

Multi-Region Consensus

ICMP can fail for purely regional reasons — a peering issue between one of our regions and your host doesn't mean the host is down. To avoid false alarms, ping monitoring uses the same multi-region consensus rule as the main HTTP uptime check:

  • An incident is opened only when a majority of monitored regions report failure (66% threshold for 5+ regions, 3 of 4, 2 of 3, etc.)
  • If even one region still reports success, no incident is opened — this prevents a single bad peering link from waking you up
  • An incident is closed only when every monitored region is back online, so a flapping host doesn't generate a recovery-then-fail notification storm

ICMP Can Be Blocked

Some cloud providers block ICMP at the firewall by default — Azure VMs, certain AWS security groups, and many corporate networks all suppress echo replies as a hardening measure. A failing ping monitor does not necessarily mean your host is down. Before treating ping failures as a real outage, verify with a manual ping from a known-good network and confirm ICMP is allowed in your host's firewall and security-group rules.

Notifications

Ping failures are treated as critical-tier incidents — the same severity as a hard HTTP outage. They route through whichever notification channels you have configured for the critical tier (email, SMS, Slack, webhook, etc.). The notification subject reads "Host Not Responding to Ping" and includes the regional context, root cause (e.g. "Timeout", "Host unreachable") and a link to the incident detail page.

When the host starts responding again from every monitored region, a recovery notification fires ("Host Responding to Ping Again") and the incident closes automatically.

Limitations

  • Private addresses are blocked — we won't ping RFC1918, loopback, link-local, or any non-public address. If you need to monitor internal infrastructure, run a self-hosted probe
  • No sub-checks except DNS and Domain Expiration — SSL and body assertions require HTTP, which Ping monitors don't perform. Use an HTTP monitor for those
  • Plan-gated — Ping is available on the Basic plan and above, same gating as DNS and Domain Expiration monitoring
API MonitoringPort Monitoring