Responding to Alerts

Alerts appear in the Alerts pane in response to a condition or some other event that requires the attention of administrators. Each alert is ranked by severity, either informational, warning, or error. The Alerts pane always shows the alerts sorted by date and time with the most recent alerts at the top. The alerts that appear in a day are separated by a thick dividing line from alerts that occur on other days.

There are a number of steps you can take when responding to alerts.

• View a description of the event or condition that triggered the alert.
• View the number of times the event or condition occurred in a predefined interval. (Alerts that occur within this interval are grouped into a single alert. For more information, see How Alert Grouping Works.)
• Assign an owner to the alert (by default, the alert is marked Unassigned).
• Change the severity of the alert if the condition or event in your organization is less or more important than the default severity.
• Diagnose and repair the condition or event that caused the alert. Detailed steps for troubleshooting the alert are displayed when you click the More button in the Alert Details dialog box.
• After the condition or event is repaired, change the alert status from Active to Resolved.
• Edit the row filter that triggered the alert to change how future alerts are displayed. For more information, see Customizing Row Filters.
• Edit the comments field to annotate the row filter.

Note

• Alerts that have their status set to Resolved are not displayed in the Alerts pane. You can view all alerts, both Active and Resolved, on the Alert History tab.

Within a few days of an initial Ultrasound deployment, you might see a flood of alerts for problems that you did not know existed. In many cases the replica sets, members, and connections might have red (critical condition) or yellow (warning condition) health ratings. We recommend the following strategy for dealing with an initial flood of alerts and health issues:

1. Review alerts regarding provider deployments to ensure that all providers were deployed successfully. Any unsuccessful deployments could cause an inaccurate picture of the overall health of the replica set. Investigate and repair provider deployment failures using the advice given in the alert and corresponding Help topic. After that, you should receive an alert that indicates that provider deployment was successful.
2. Use the Summary tab to investigate any objects that are red (critical condition) or yellow (warning condition) and repair those issues. In many cases a corresponding alert will describe the problem. See Understanding Health Ratings for more information.
3. Resolve all alerts by using the procedure To respond to an alert, and then wait for new alerts to occur. If these alerts are informational, address them when you have time. Many informational alerts can be safely resolved without following the advice in the alert, though in some cases following the advice in an informational alert could result in better performance and reliability of your replica sets.
4. If you continue receiving alerts or health ratings that are not applicable to your organization, you can customize or disable them by using the procedure To respond to an alert.