Environment

• YugabyteDB Anywhere Platform - 2.12+

Issue

When Alerting is configured in the YugabyteDB Anywhere platform, the platform sends frequent health reports. A high volume of health reports may make it difficult to notice a health check failure.

Resolution

Overview

The following clarifies the purpose of several fields in the YugabyteDB Anywhere platform Health & Alerting configuration:

• Health check interval is how often to perform the health check (in minutes). The default Health check interval is 5 minutes.
• Active alert notification interval field defines the notification period (in minutes) for resending notifications for active alerts. The default value of 0 means that only one notification is issued for an active alert.
• Health Check email report interval is how often (in minutes) the platform should send scheduled reports about the health of the cluster. These scheduled updates always include the status of all the nodes, even if the cluster is healthy. The default Report email interval is 720 minutes.
• The Only include errors in alert emails toggle controls whether health check messages are sent for healthy clusters or if they are sent only when there is an issue.

Setting the  email report interval to a very high value will suppress reports for healthy universes.

Steps

1. Open the Admin settings, from Alert Configurations select the Health tab:

2. Set the Callhome Level to Low, Medium or High.

3. Set the Health check interval to the desired alerting interval. For example, if this interval is set to 5 minutes, the health check will run every 5 minutes. A health check message will be sent if there is an issue in the cluster or if the Only include errors in alert e-mails toggle is disabled.

4. To send email notifications for health checks, turn on the Email notifications option and enter a comma-separated list of email addresses to which alerts are to be sent in the Alert emails field.

5. Set the Health Check email report interval to a very large number (e.g. 525600 minutes, which is 1  year). This will suppress scheduled reports on cluster health.

6. Enable the Only include errors in alert e-mails toggle. This will prevent health check messages from being sent out if the cluster is healthy.