SQL Monitor 2.3

Walkthrough 1: General overview

SQL Monitor is a multi-server performance monitoring and alerting tool for SQL Server. The UI is web-based, so it can be accessed from desktop machines and mobile devices. It has four tabs: Overviews, Alerts, Analysis and Configuration.

Above, we have the Overviews tab. This allows us to view the current status of any part of our database system. We can take a global view, as above, or drill down to the cluster, machine, SQL Server or database level.

The Global Overview shows us the current alert status of each machine (High, Medium or Low), the number of uncleared alerts, the Processor Time and Memory used.

For more detail, we can select a particular machine or SQL Server, using the Monitored servers list on the left hand side. Let's take a look at our itanium2 machine.

Here's the overview for itanium-2. We can see the SQL Server instances on the machine, the alert status and number of alerts for the machine and SQL Server, and various metrics. The metrics are updated in real time, so we become aware of problems as soon as they arise. Clicking on a sparkline will open the Analysis tab with a more detailed view of the data for that particular metric. For more information see the Working with Analysis walkthrough.

If we wanted to drill further down, we could click on the SQL Server instance – named (local) – in either the main view or the Monitored servers menu. For more information, see the Working with Overviews walkthrough.

Next, we'll move to the Alerts tab.

Clicking on the Alerts tab takes us to the Alert Inbox. Whenever a problem arises, a message appears here, indicating the severity of the problem (through the Alert Level), the object affected, and the time the problem occurred.

The Alert Inbox is accessible by everyone who has been given the SQL Monitor password, and we can mark alerts as cleared or uncleared, read or unread, as we deal with them. If we want to include more information, we can also leave comments. Clicking on an alert takes us to the Alert Details page.

On this Alert Details page we have the information for a blocked process. We can see details of when and where our databases were affected, as well as some more detailed performance information. For example, we can investigate how processor time was affected on our Host Machine when this happened.

For more details, see the Working with Alerts walkthrough. For now, we're going to move on to the Analysis tab.

The Analysis tab is used for reporting and provides a graphical overview of the data SQL Monitor has gathered. We can choose the metric to look at, and specify each level to which it applies (e.g. which SQL Server on which machine). Once we've chosen a metric, we can specify the time range with the drop down menu. On the right, we also have a description of the metric. Its maximum, minimum and mean values are viewable in the Statistics tab.

From the drop down menu, we can also choose to compare the metric over two equivalent time periods. They display on the same graph, for ease of comparison, and our statistics update to include both. For more details, see the Working with Analysis walkthrough.

In addition to looking at historical data in the Analysis tab, we can also go back to a particular point in time from the clock in the top right corner on the overviews. We can pick a particular date, or enter a specific time and date in the dialog boxes to the right. When we click Apply, we enter Back in Time mode.

Back in Time mode presents us with static data, fixed at the time we chose. We can use any of the overviews to see what was going on. We can also navigate forwards and backwards in time, using the buttons at the top. By clicking on a particular metric, we can move straight to the Analysis tab. For more detail, see the Rewind Time walkthrough.

Finally, under the Configuration tab, we can adjust various settings, such as our monitored servers list, groups of servers, email settings, and so on. As an example, here's the list of Alert settings. To configure an alert, we just click on it.

This takes us to an Alert Configuration, where we can specify Alert level thresholds, particular objects or T-SQL commands to exclude for the long-running query alert, and which address to email when the alert occurs. The Alert Configuration settings are also available directly from the Alerts tab. For more information, see the Working with Alerts walkthrough.