Viewing alert details

Click on an alert to view more information about it in the alert details pane below the list of raised alerts:

The alert details pane shows a summary of the alert, and provides a number of expandable sections. These sections contain further information about the alert, and diagnostic data, to help you identify what was happening on the SQL Server at the time the alert was raised. Click on a section heading to expand or collapse that section.

The first expandable section is specific to the type of alert raised. For job-related alerts, for example, the Job step details section lists the steps of the job, and if appropriate, the outcome of each step. For Long running query, Blocked SQL process, and SQL deadlock alerts, the first section lists the processes involved and includes a query fragment for each process.

All data for the alert is relative to the time the alert was raised. For example, the Next scheduled to run time for job type alerts shows the next run after the displayed alert time; if you are viewing an alert raised several hours or days previously, the time of the next run may be in the past.

You can adjust the height of the alert details pane by dragging it up or down, or by using the buttons to move the pane to a fixed height.

Click to open the alert in a new window.

Time zones of alert times

All times shown in SQL Response are local to the SQL Server on which the alert was raised.

For example, if you are monitoring SQL Servers in Moscow on a computer running SQL Response in New York, then the alert times displayed in SQL Response will all be eight hours ahead. If you are monitoring SQL Servers across numerous time zones, this means that sorting the list of alerts by the Last occurred column may not give you a chronological list of the most recent alerts.

Viewing alert occurrences

When you view an alert, the latest (most recent) occurrence is displayed. You can review earlier occurrences of the alert by doing one of the following:

• click the previous occurrence button
• from the Viewing this occurrence list, select the occurrence you want to view.

  

  The most recent occurrences are listed first, but you can scroll down to view occurrences not within the current timeframe (shown in grey).

How alerts are grouped

A new occurrence of an alert is raised when the same problem happens again on a SQL Server; that is, the basic details of the alert are identical. For example, when the same job fails twice in an hour, an additional occurrence of the Job failure alert is raised for that specific job; the job name is shown in the Detail column. If a different job fails, a separate Job failure alert is raised.

• For connection type alerts (SQL Server unreachable, Computer unreachable, SQL Server login failure, and Computer login failure), a second occurrence of the alert is raised only when the connection has first been re-established and then fails for a second time. If the SQL Server remains offline or unavailable, only the initial occurrence of the alert will be listed.
• For SQL deadlock, Blocked SQL process and Long running query alerts, repeated instances of the problem are grouped into a single alert, rather than into separate alerts for each process or query. Select an occurrence of the alert to view the details for the specific problem.

  For example, a number of different queries overrun on a SQL Server within a half-hour period, resulting in 20 occurrences of a single Long running query alert, rather than 20 separate alerts in the main list. Use the Viewing this occurence list in the alert details pane to review the query involved in each separate overrun.

When you clear an alert, you clear the status of the alert as a whole; you cannot clear an individual occurrence of the alert. Clearing an alert identifies the current status of the problem rather than being a property of an individual occurrence. See Clearing alerts and adding comments.

Viewing the performance snapshot

The Performance snapshot section provides a snapshot of SQL Server and CPU activity leading up to the time the alert was raised, and for a short period afterwards. As SQL Response is monitoring your SQL Servers continuously, this data is already captured, and can be included in the details of the alert. If you have enabled the collection of SQL Profiler trace on a monitored SQL Server, then you can also view the Transact-SQL statements for each connection to that SQL Server in the period leading up to the alert.

See Viewing performance snapshot data.

Performance snapshot information is available for all alerts except the following:

• Computer login failure, SQL Server login failure
• Computer unreachable, SQL Server unreachable
• SQL Agent not running

Viewing system processes

The System processes section lists the Windows processes that were running at the time the alert was raised.

For each process, the following data is provided:

• ID: The numerical ID assigned to the process by Windows while it was running.
• Start Time: When the process started.
• Approx End Time: When the process ended. This column is empty when the process was still running at the time the alert was raised.
• CPU Time: The total processor time, in seconds, used by the process since it was started, at the time the alert was raised.
• Mem Usage: The amount of main memory used by the process, at the time the alert was raised
• VM Size: The amount of virtual memory, or address space, committed to a process, at the time the alert was raised.

These performance measures are the same as those available in Windows Task Manager.

Click on a column heading to sort the processes by that column.

The data provided is a snapshot of Windows process information from around the time the alert was raised. This data could be from up to ten seconds before the exact time the alert was raised; process data is captured every ten seconds.

Viewing comments

Comments can be added to an alert only when that alert is cleared; they allow you to add notes about the causes of the problem and describe any actions taken to resolve the issue. When comments are added to an alert, the time and user name are added automatically.

Comments are not attached to any specific occurrence of an alert. If an alert is cleared, re-raised, and then cleared again several times, each additional comment will be appended to the list. All comments for an alert are always available to view.

To view the comments made on alerts that are currently at Cleared status, ensure that the Cleared alert filter check mark is selected in the Alert levels to show area of the Filter pane.

See Clearing alerts and adding comments.

Selecting multiple alerts

You can select several alerts in the main list at once.

• To open each selected alert in a new window, click
• Click Clear Selected Alerts to clear multiple alerts at once. Adding a comment to the cleared alerts will add the same comment to each alert.

Viewing recommendation details

Recommendation details do not contain multiple sections. Each recommendation contains basic information about the data that caused the recommendation to be raised, and a description of the issue.

See About alerts and recommendations.