The Overview tab is displayed by default. As indicated in the name, this view is showing the overall health, and the 10 most recent alerts that fall within the selected time frame (within 24 hours) for all monitored targets in your environment.
Health Score Calculation
The Health Score (shown above) is calculated by incorporating open alerts and their associated severity. It uses a 24-hour window in the calculation, and can be adjusted with the Time Slicer to display more recent alerts in varying hour and minute increments. A high health score (with a maximum of 100 possible) is an indicator of good health.
The points per severity level are:
- Low - 1 point
- Medium - 3 points
- High - 6 points
- Critical - 9 points
Note: The weight of an open alert diminishes with its age. The score being presented is 100 minus the sum of the weighted value of open alerts. Closed alerts are not factored into the score. Alerts can be closed in the Events Log in the SQL Sentry configuration client.
Select a time increment from the Time Slicer to adjust the Health Score, wedgets, and associated alerts based on your selection:
Select the (chevron-right icon) to the right of the count to display the individual details about each logged alert on any of the health views:
EHO for Sites and Groups
The Environment Health Overview (EHO) is available for sites and groups in SQL Sentry Portal. Use the Show Hierarchy switch to toggle the list of targets between a flat list and a list grouped by sites.
Expand a site to view the groups within it and select a group to view a group EHO.
Note: The Site and Group icons and labels note which type of overview you are viewing.
All SQL Servers View
Select All SQL Servers from the sidebar to display the overall health view for all SQL Servers in your monitored environment.
Note: Uptime is a measure of the percentage of time that SQL Sentry Portal is able to connect to the server to collect data.
The All SQL Servers view provides options for viewing overall health through Alerts by Severity, Alerts by Tag, and Wait Time / Session. These views are similar to the target health views described below, but they include the overview information for all monitored SQL Servers.
All Windows Servers View
The All Windows Servers view provides options for viewing overall health through Alerts by Severity, Alerts by Tag, and Wait Time / Session. These views are similar to the target health views described below, but they include the overview information for all monitored Windows Servers.
Target Health View
When selecting a target, the Health view is the default view. This screen displays wedgets for Alerts by Severity, Alerts by Tag, and Wait Time / Session. The default view is Alerts by Severity.
Alerts by Severity
In the example below, the Alerts by Severity wedget is displaying 5 Critical and 9 Medium severity alerts. In the Alerts table, you can see 2 rows of alerts with a severity of critical (4, and 1) totaling 5, and 8 rows of medium alerts (with one row displaying 2 alerts) totaling 9. This is how the number in the color-coded alert is calculated.
Note: Selecting the blue hyperlinked target name on the left provides an option to switch to the Health, Performance, Storage, Top SQL, Blocking, Deadlocks, or TempDB view for that target.
Alerts by Tag
The Alerts by Tag view works by incorporating alerts to calculate the scores and associate them to tags on the alerts (Network, CPU, Memory, Disk, and Other). The score for these views is based on the severity of all open alerts within the selected time period. A high health score (with a maximum of 100 possible) is an indicator of good health.
Wait Time / Session
Wait Time / Session relies on wait stats data collected by SQL Sentry. These wait stats are broken down by major resource category (Network, CPU, Memory, Disk, Other). In this view, the health score worsens as waits increase. A low ms/session score is an indicator of good health.
Each category is calculated independently. The displayed value for a category is calculated as the waits for the category divided by the user sessions. The values for waits and user sessions are based on the most recent values present in the selected time frame. Each value is then rounded and displayed as the category’s value. The overall score is the sum of the five rounded categories.
The Other category is for other important wait types that either affect performance in more than one major category, or can't be directly attributed to any category with absolute certainty, such as backups and parallelism respectively.
As seen in the image above, there is a list of Related Targets (1 SQL Server and 1 Windows Server in this example). When there are multiple instances on a SQL Server, they will be listed as a related target, as will the Windows Server where the SQL Server is installed.