How an investigation starts
The watch service checks your pipeline every 30 seconds. A single bad check is not enough to open an investigation.
AvailabilityAvailable to everyone with early access to Varqeni.
Steps
- When two checks in a row fail, the watch service opens an investigation.
- While that fault lasts, it does not open another one for it.
- Once the pipeline has been healthy for the same number of checks, the incident is over. A new fault after that opens a new investigation straight away.
- A notice pops up at the bottom of the console: The agent opened an investigation. Choose it to open the investigation.
What you should see: The new investigation at the top of the Investigations list, with its cause.
What it can and cannot do
Can
- Open one investigation per fault after two bad checks in a row, and a new one at once if the fault returns.
Cannot
- Open an investigation for a single bad check, or repeat one while the same fault lasts.
Related articles
- Read an investigationOpen an investigation to see the cause of record, the model review, what was ruled out and what a person should do next.
- Filter the investigations listShow all investigations, only faults, only those the watch service opened, or blind trials.
- Notices in Microsoft TeamsVarqeni posts a card to a Teams channel when it opens an incident, when the autopilot acts, and when a downstream reader needs a person.
Still stuck? Contact support or email info@fortisdatagroup.com.