Technical Requirements
dbWatch is packaged for Windows and Linux platforms and only officially supported for these platforms; however, dbWatch is Java based and will work on any platform that supports Java.[…]
dbWatch Server
- Windows or Linux Server (VMWare virtual server supported)
- 1 GB RAM
- 1 GB HD Space
- Installs in under 15 minutes
dbWatch Engine
- 100 Mb free space in database instance
- Engine – Server communication determined by each supported database platform.
- Bulk install for large database environments
- Installs in under 2 minutes per instance
dbWatch Client
- Windows or Linux operating system for use with graphical interface
- 500 Mb hard drive space
- Mouse
- Java support
- Client – Server communication requires single port only
- Client is installed automatically under Server installation
Using dbWatch
When you start a fresh dbWatch installation for the first time, the “Add database” wizard shows up and guides you through the process of installing the dbWatch framework in its own database on a SQL Server instance. This database hosts all the internal database objects that dbWatch needs in order to fulfil its job. Nothing is installed in any of SQL Server’s system databases.
On the first page of the wizard, you select the database platform. Furthermore you select the type of installation. This type determines whether you will install a default set of tasks, alerts, and configuration settings; or rather choose your own subset of these. But since you can customise everything later on to suit your own needs, we just go for now with an “Express installation” and demonstrate how to change that later on.
dbWatch’s monitoring databases are organised in groups. By default there are 3 groups available right from the start:
- Development
- Test
- Production
These are “usually” the different environments you have, so these groups make sense. However, if you want to have your own group, you can do so and I’ll show you later on how to create new groups. For now, I just give the new connection a name and assign it to the Development group.
Next step in the wizard is to provide your connection credentials. You have the choice between choosing to use Windows Authentication or SQL Server Authentication. I have chosen to use SQL Server Authentication.
Then you supply details for the monitoring database such as database name and the login that dbWatch will use to connect and that will be in the dbo role for that database.
If you click on the “Details” button in step 3 above, you’ll come to a dialog where you can set more options for this new database. As you can see, you can change the physical location and the size of both the data and the log file. For my purposes I just accept the defaults and just close the dialog again.
On the second but last page you have the opportunity to decide on which predefined tasks and checks you want to install in the new database. Just take with you that you can change this later on at any time. I’ll come back to it when I describe some of these tasks more detailed.
The last page of the wizard offers you a summary of the tasks that will be carried out and as soon as you click on “Finish”.
The above screenshot of the main window of dbWatch after the wizard has finished. As you can see, the GUI is well-arranged and not overloaded. Let us now come back to a step I skipped before and have a look at the tasks and checks that have been installed by the wizard.
When I expand the tree for the “SQLServerConnection” that I’ve just created I can see all the alerts and tasks available to this connection. A click on any item in the navigational tree view below the connection node is synchronised to the right-hand detail pane and the corresponding row is highlighted.
In the above example, I have selected the “Instance memory check”. Let’s take this alert and look at it a bit more closely. A quick overview over its status is given in the right-hand detail window. There you can see information like “Last Run”, “Next Run”, and the “Schedule”.
A double-click on the row opens a message box that also gives you essential information about this alert. This is certainly a useful feature for a quick ad-hoc overview in situations where you do not need a full-blown report.
Through the context-menu either of the tree view or the main detail pane you’re in full control of the most important tasks pertaining to this alert, such as running the job manually out of the scheduled cycle, configure, disable and enable, or edit it. Let’s look at the different option one by one.
