Dial testing — endpoint availability

Dial testing — endpoint availability

Active monitoring of network endpoints: the platform dials a target — HTTP, TCP, or ICMP — on a schedule and records whether it responded, from the agents’ vantage point. This is the availability-monitoring surface, reached as Dial Test in the collection area.

---

What it does

A dial rule defines a target and a check; agents distributed to the rule run the check on the configured frequency and report pass/fail plus response details. This answers “is the endpoint up and fast from where our agents run” — different from log/metric collection, which observes what already happened.

type	what it checks
HTTP	a URL returns the expected status code and/or text — the web endpoint check
TCP	a host:port accepts a connection — the service check
ICMP	host reachability via echo requests — the network check (confirms the host is up, not that any service on it is)

Creating a rule

Collection → Dial Test → Create, with three panels:

1. Select Dial Test Type — HTTP, TCP or ICMP.
2. Basic Configuration — name, the target (URL for HTTP, host:port for TCP, host for ICMP; multiple comma-separated addresses allowed), and the dial frequency (interval, or a cron expression with the standard none/skip/delay overlap policy).
3. Advanced Configuration — per type: request settings (method, headers, body), response checks (expected status code / text / content matching), TLS/SSL settings for HTTP and TCP.

Running and distributing

A rule is run by agents: after creation, Distribute Agents assigns it to specific agents or agent tags. The rule list shows name, hosts, type, status, update time, and per-run Test Result (succeeded/failed with the error message). Rules can be enabled, disabled, edited, and deleted. A Test action runs the check once immediately to validate the rule before distribution.