How can we help?

Component Status Monitoring - Best Practices for Success

Follow

Auvik’s alerting is flexible to identify common elements on a single network device entity. For example a switch stack can have multiple fan components. How alerting can work out for you is largely contingent on two things.

ANY Alert triggers when any sub-item on the device meets the trigger conditions.

Example: If a component status EQUALS bad

  • If multiple component statuses share the same status, they will all get lumped into the same alert.

ALL interfaces block

  • Alert triggers only if every interface matches the rule.

    Example: If all interfaces have admin state down → alert.

The “empty” status gotcha

Quite often, a rule like “Component status does not equal GOOD” is sufficient. However “empty” is a valid state, and exist for expansion slots that may be unused. An alert like this will forever be in a state of triggered. If not cleared, it will prevent future alerts.

Users should create a combination of rules such as:

If any component matches ALL of the following conditions:

Component State is NOT GOOD
AND
Component State is NOT EMPTY

Create Alerts based on Component Sensor Data

Users can create alerts for devices that have component sensor data displayed on the device details page. These will show up under the component sensor category.

To create an alert, select a trigger from the drop down:

definealertrule.png

It is recommended for component alerting to use the For ANY component setting. To reduce noise, use the Roll up Alert function. This will create only a single alert for all sensors that may trigger - which is quite likely for component sensors Build alerts by adding trigger condition logic.

definealert.png
Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request