Integration with GLPI

This integration turns Mugnsoft monitor status changes into GLPI tickets. When a monitor breaches, the Integrator opens a ticket; when the monitor recovers, the same ticket is resolved automatically.

As a result, your IT service desk sees Mugnsoft outages as first-class tickets, with no manual ticket creation or closure.

Note:

This integration is configured on the Integrator component, alongside the other push targets. It uses the GLPI REST API over HTTPS, so no GLPI plugin is required.

We assume there are no firewall rules preventing the Mugnsoft Integrator from reaching your GLPI instance, and that you have an integration user with permission to create and update tickets.

How it works

The Integrator reacts only to a confirmed status change:

Breach (monitor goes to a non-OK status): a new ticket is created with a title, the monitor detail in the content, and urgency/impact derived from the severity.
Recovery (monitor returns to OK): the open ticket is found again and transitioned to its solved status.

GLPI has no native correlation field, so each ticket carries an embedded key in its title (built from the monitor type, name, and probe). The open ticket is located through the GLPI search API on that key. Because the lookup is server-side, resolve still works after an Integrator restart, and repeated non-OK checks never open duplicate tickets.

Mugnsoft status	GLPI urgency / impact
`CRITICAL` / `ERROR`	`5` (very high)
`MAJOR`	`4` (high)
`MINOR`	`3` (medium)

GLPI derives the ticket priority from the urgency × impact matrix.

Acknowledgement and ticket ownership

Mugnsoft itself has no acknowledgement concept — alerting is purely status-driven. To avoid auto-resolve closing a ticket your team is actively working, the integration treats the GLPI ticket status as the acknowledgement signal:

If, on recovery, the ticket is still New, it is auto-resolved.
If a human has taken it (assigned, processing, or pending), the ticket is left open and a recovery follow-up is posted instead — closure stays with the operator.

Set Auto-resolve off to disable automated closure entirely (a recovery follow-up is always posted).

Flap dampening (grace period)

Because alerting is status-driven and has no dampening of its own, a flapping monitor could open and resolve a stream of tickets. The Grace period delays opening a ticket after a breach: if the monitor recovers within the window, the flap is suppressed and no ticket is created. A genuinely sustained breach still opens a ticket once the window elapses. Set it to 0 (default) to open immediately.

Prepare GLPI

In Setup > General > API, enable the REST API and note the API base (the apirest.php endpoint is reached over your instance host).
Create (or choose) an API client. If the client is configured to require an App-Token, copy it — you will paste it into the Integrator. If no App-Token is required, you can leave that field empty.
Create (or choose) an integration user with a profile allowing ticket create and update (for example Technician).

Configure the Integrator

Navigate to the Integrator component settings and open the external services panel.
Select the GLPI tile and enable it.

GLPI integration configuration on the Integrator

Fill in the connection fields:

Field	Value
IP/FQDN	GLPI instance host (e.g. `glpi.example.com`)
Port	`443`
SSL	on
User	the integration user
Password	the integration user’s password
App-Token	the API client App-Token (leave empty if the client has none)
Solved status	ticket status used on auto-resolve (default `5` = Solved; `6` = Closed)
Auto-resolve	on (default). When a human has taken the ticket, it is left open and a recovery follow-up is posted instead of closing
Grace period (s)	flap-dampening delay before opening on breach; recovery within the window creates no ticket (default `0` = immediate)

Click the test button to verify connectivity and credentials, then save.

Verify

Force a monitor failure (for example, run a URL monitor against an unreachable target). A new ticket should appear in GLPI with the monitor name and an embedded key in the title.
Restore the target so the monitor returns to OK. The same ticket should transition to its solved status — no duplicate, and no second ticket on subsequent OK checks.