Quick send to 3rd parties
Results leave Mugnsoft through the Integrator. This page turns on the two things it can do with them: notify people and feed your other tools.
How results travel
- A probe runs a check and sends the result to its Integrator.
- The Integrator keeps it as an event, sends the notifications, and forwards the data to the tools you switched on.
A probe reports through the Webserver’s Integrator unless you give it another one, so with one Integrator there is nothing to link.
Step 1: open the Integrator settings
Open gear menu → components, right-click the Integrator and choose edit server.
Step 2: switch on a tool
In the External Integration tab, each tool is a tile.
- Click the tile of your tool (Splunk, Elastic, InfluxDB, Kafka, Zabbix…).
- Switch it on and fill in its connection.
- Test it with ⚡.
- Save with the disk icon.
Integration has one page per tool, with the fields to fill in and what arrives on the other side.
Step 3: switch on notifications
Email, Slack, Teams and PagerDuty are set in two places:
| Where | What you set |
|---|---|
| The Integrator | The channel itself and its default destination (mail recipients, Slack channel, Teams webhook…). |
| Each monitor, tab Actions | Which channels this monitor uses and on which statuses. Leave the destination blank to use the Integrator’s default. |
Email also needs the mail server: set it once in the Webserver settings, SMTP tile, and turn the tile’s switch on.
Step 4: check that data flows
- Open Events (the list icon): every status change that reached the Integrator is listed.
- For a local copy of every result, switch on log data events to CSV file in the Integrator’s General tab. The files are written in the Integrator’s data directory.
Several Integrators
To send a probe’s results to a different Integrator, or to a second one for resilience: right-click the probe, choose edit server, and add it to Integrator endpoints under Integrators & notification resilience.
Next
- Integration: the detail for each tool.
- Integrator: buffering, retries and what happens when a tool is down.
- Downtime operations: silence alerts during maintenance.