In SQL Diagnostic Manager, whenever a metric exceeds the configured value or threshold, the state of the metric changes and it can execute an alert or action. In SQL Diagnostic Manager, thresholds come pre-configured with industry-standard levels. This does not mean these levels are a one-fit-for-all solution for everyone. There are several attributes regarding thresholds in SQL Diagnostic Manager. First, when you select the instance configure alerts, you can have the option for configuring the thresholds. You can view the metrics that you can enable or disable, meaning that you can collect information about those metrics. You need not collect data you do not need or you do not want. The next feature is ranking, which is new in SQL Diagnostic Manager 10.5. It is basically an alert priority condition to only notify you of the most important alerts above a certain track. By default, they come with certain ranks. When you create an alert, you can specify that it only alerts on those metrics that rank above a certain threshold.
Watch the 18-minute video “Configuring Metric Thresholds and Baselines with SQL Diagnostic Manager for SQL Server” to learn how to better alert to abnormalities while not alerting to normal situations for SQL Servers by quickly and easily configuring metric thresholds and baselines.
With SQL Diagnostic Manager for SQL Server, configure alerts to inform and warn about approaching issues with SQL Server instances. View these alerts using the desktop, web, and mobile console, and the newsfeed. When reaching alert thresholds, send email notifications, pop up alert messages in the Windows taskbar, write events to the Windows Event log, generate events on the timeline, and send alert messages to the newsfeed action provider. After correcting problems triggering alerts, alert again when situations recur.
Choose from over 100 pre-defined and configurable alert settings based on industry best practices with advanced configuration settings to allow for greater flexibility. Configure general alert settings as templates to apply to other servers and groups of servers. Set different alert thresholds for each database or disk within each monitored instance. Capture baselines of past performance of monitored instances to determine whether alert thresholds are excessively noisy or cause a false positive. When metrics are continually alerting, view flags to show that changes may be necessary and see recommendations for new limits.
Easily snooze alerts and groups of alerts for a specified time to prevent alerts from recurring while resolving problems. Determine the time that an issue must occur before sending first alerts for problematic metrics that occasionally spike for very short durations. Disable data collection and alerting during maintenance periods to avoid false positives. Maintenance mode can be on-demand, onetime, and or weekly scheduled maintenance periods. Enable maintenance mode via PowerShell to integrate with an outside job scheduling.