Probe locations
A monitor’s locations is the set of places it is probed from. Each one probes independently on the
monitor’s own interval, and each one carries its own up/down verdict; the monitor’s verdict is the
quorum over them.
A place a probe can originate from is a probe_location. It is one of two things:
- a WireWatch region — infrastructure we run, shared by every organisation;
- one of your agents — a WireWatch binary you run inside your own network, serving only your organisation.
The set is open
Section titled “The set is open”The set of probe locations is not fixed. It is the WireWatch regions plus your organisation’s
active agents, and both halves move: a region is added by deploying one, an agent is added by
enrolling one and disappears when you revoke it. Neither is an enumerated type anywhere in the
system — locations is stored as open text, and adding a region needs no schema change and no
release on your side.
So do not hard-code the table below. The list your organisation can actually pick from, right now, is
GET /orgs/:orgId/probe-regionswhich is what the location picker in the dashboard is reading, plus your agents from the agent list.
The regions today
Section titled “The regions today”As of this writing WireWatch probes from three regions:
| Id | Location |
|---|---|
us-east |
US East (N. Virginia) |
eu-west |
Europe (West) |
br-east |
Brazil (São Paulo) |
The id is the contract; the label is for humans. Ids are provider-neutral geographic names on purpose — they name where the vantage point is, never who hosts it, so moving a region between providers changes nothing you have configured.
Your own agents
Section titled “Your own agents”A region cannot reach a target inside your network, and it should not be able to. An agent is how a private target gets monitored: you run it, it authenticates to WireWatch with a client certificate, and it serves exactly one organisation.
An agent appears as a probe location under its own id. Where a region is a name like us-east,
an agent is the UUID it was enrolled with, and it goes into locations the same way.
Two consequences worth knowing before you configure one:
- A monitor whose locations are all agents may target a private address. That is the entire point of an agent, and it is the only way the reachability guard is relaxed. Mix in a single region and the target must be publicly reachable again — because that region would have to dial it.
- An agent id that is not one of your organisation’s active agents is refused, and refused identically whether it never existed, was revoked, or belongs to somebody else. A monitor cannot be used to discover another tenant’s agents.
Agents are not available on every plan — agents is a quota, and it is 0 on Free.
How many locations you may use
Section titled “How many locations you may use”probe_locations is a plan quota, and it bounds the size of one monitor’s locations set:
| Plan | probe_locations |
|---|---|
free |
1 |
basic |
2 |
premium |
3 |
enterprise |
5 |
Exceeding it is refused with 402 and the metric named.
Why more than one location changes the answer
Section titled “Why more than one location changes the answer”With a single location, a failure is ambiguous: the target may be down, or the path from that one vantage point may be. WireWatch resolves the ambiguity by asking a different location before opening an incident — see Getting started for the sequence — so even a single-location monitor gets a second opinion.
With several, the disagreement is the signal rather than a tie-break: all locations failing is a different incident from one location failing, and they are recorded with different severities. If you only ever raise one number, raise this one.