How do I manage device health check frequencies?

Setting the health check frequency for your devices controls how often Auvik checks whether they are online.

How does a health check work?

When the collector performs a health check, it sends multiple types of ping requests to the device. If the device responds to any of them, the health check is successful. If the device does not respond within the required time, the health check fails.

Auvik uses consecutive health-check failures to determine when a device should be considered offline. By default, Auvik waits for five consecutive failures before marking a device offline.

If a device fails a health check but then responds successfully to an SNMP poll before the failure threshold is reached, the failure count resets to zero.

For example, if an alert requires three consecutive failures:

The first health check fails.
The second health check fails.
A successful SNMP poll resets the failure count.
The first health check fails again.
The second health check fails again.
The third consecutive health check fails and the alert condition is met.

How do frequency and consecutive failures affect alert timing?

Two settings primarily determine how quickly an offline condition can be detected:

Health check frequency: How often Auvik checks the device.
Consecutive failures: How many failed checks are required before the device is considered offline.

A rough estimate for the earliest detection time is:

Health-check interval × the number of intervals needed to reach the failure threshold

A small processing delay may occur after the threshold is reached.

For example:

With a 15-second health-check interval and five required failures, detection may take approximately 60–75 seconds.
With a one-minute health-check interval and five required failures, detection may take approximately four to five minutes.

The exact timing can vary depending on when the first failed check occurs, health-check scheduling, collector workload, and whether a successful SNMP poll resets the failure count.

How do Alert v2 delays affect notification timing?

For Alerts v2, the device may be identified as offline before an alert notification is sent. If the alert has a trigger delay, the offline condition must remain active for that delay before the alert fires.

The approximate timing is:

Health-check detection time + Alert v2 trigger delay + processing time

For example, if a device is identified as offline after approximately one minute and the Alerts v2 definition has a five-minute trigger delay, the notification may be sent approximately six minutes after the initial outage.

A trigger delay can reduce notifications caused by short interruptions, but it also delays notification of genuine outages.

When are email notifications sent?

Email notifications are sent after the alert is created and the notification policy processes it.

Email delivery can take slightly longer because of:

Email frequency or batching settings.
Notification suppression rules.
Maintenance windows.
Alert deduplication or repetitive-alert controls.
Mail-provider routing and filtering.

If an alert appears in Auvik but the email arrives later, review the notification policy and email frequency settings before assuming that the health check is delayed.

Health-check frequencies

Screenshot 2024-05-08 at 11.54.23 AM.png

Auvik provides preconfigured health checks for common device categories, including:

Access Points.
Network Elements.
Default Health Check.

You can edit, delete, or change the priority of these health checks. You can also create custom health checks for specific device types or groups.

Health checks can define:

The devices or device types being monitored.
How often devices are checked.
The number of failed checks required before a device is considered offline.
The priority in which health checks are run.

Important: Health-check priority affects which checks are executed. If a higher-priority health check matches a device, a lower-priority matching health check may not run.

How to add a health check

Adding a health check from the MSP dashboard applies it to all clients. To add one for a specific client, add it from that client’s dashboard.

Click Discovery in the side navigation bar.
Click the Discovery Settings tab.
Click Health Check Frequencies.
Click Add Health Check Frequency.
Complete the form:
- Select whether the check applies to all devices or a specific device type.
- Select how frequently devices should be checked.
- Specify how many failed checks are required before a device is considered offline.
Click Save.

After adding the health check, review its priority to ensure it runs as expected.

How to edit a health check

Click Discovery in the side navigation bar.
Click the Discovery Settings tab.
Click Health Check Frequencies.
Find the health check you want to edit.
Select the checkbox at the beginning of the row.
Click Edit.
Make the required changes.
Click Save.

Changes made from the MSP dashboard apply to all clients. To change a health check for one client, make the change from that client’s dashboard.

How to reset a health check

Only health checks preconfigured by Auvik or created from the MSP dashboard can be reset.

Click Discovery in the side navigation bar.
Click the Discovery Settings tab.
Click Health Check Frequencies.
Select the health check you want to reset.
Click Reset.
Confirm the reset.

How to delete a health check

Click Discovery in the side navigation bar.
Click the Discovery Settings tab.
Click Health Check Frequencies.
Select the health check you want to delete.
Click Delete.
Confirm the deletion.

Deleting a health check from the MSP dashboard deletes it for all clients. To delete one for a specific client, delete it from that client’s dashboard.

How to change health-check priority

Click Discovery in the side navigation bar.
Click the Discovery Settings tab.
Click Health Check Frequencies.
Find the health check you want to reprioritize.
Click the up arrow to increase its priority or the down arrow to decrease it.

Lower numbers indicate higher priority.

How to estimate why an email arrived late

When investigating delayed offline notifications:

Identify the health-check frequency for the affected device.
Identify the number of consecutive failures required.
Estimate the detection time using the health-check interval and failure threshold.
Add any Alerts v2 trigger delay.
Review email frequency, batching, suppression, maintenance windows, and deduplication settings.
Allow for normal email-provider delivery time.

If the device recovers before the required consecutive-failure threshold is reached, the offline condition may never be created.

How can we help?