Using Qure Analyzer
When you start up the Qure Analyzer, you will see the above depicted simple and clean main interface. There is no guesswork as where to find things. You can either create a new workload analysis, or open an existing one. For the purpose of this review, I am not going to create my own workload analysis, but rather focus on the demo analysis. This is to make the review “less subjective” and the numbers presented “more reproducible”. But before this, let me just show you how easy it is to create a new analysis:
By clicking on the “New Workload Analysis…” link you open the dialog above. As you can see, you have the choice of saving the raw data for the analysis either to file or to a SQL Server database. Each option has its own advantages and disadvantages. Saving to file offers greater portability, which might be useful when you collect data at a client’s side for later analysis in your office, for example. However, the maximum file size is capped at 4 GB. Saving to an SQL Server database on the other hand offers better performance and does not limit you to 4 GB. Of course you lose a fair chunk of the portability of the file approach. Also, the comparison of workloads is only available when the analysis is saved to a database. But more on this feature later on.
After opening the demo analysis, you arrive on the “Summary” page on a new tab. This summary gives you a high-level overview of the analysis, such as number of event and resource statistics. From there you can drill down to any deeper level of analysis. This can either be done by following the links in the “More information” column of any grid or by clicking on the details tab.
Both ways described above bring you to the “Details” tab where you can see the interface into the multi-dimensional analysis capabilities of the Qure Analyzer. This is also the heart of the Analyzer.
Let’s have a closer look at this screen.
The left-hand side gives you full control on the grouping level of the right-hand grid. You can have many groupings over multiple columns and order and nest them in any way you like. Once you are happy with it, click “Apply” to apply them to the result grid.
In the same manner by which you define grouping levels, you can apply multiple filters to your workload to narrow down your analysis. There are three different types of filters available in the Qure Analyzer: Text Search, Discrete Values and Sliding Range Controls:
The “Text Search” feature allows you to specify a search string that must be contained in the batch name or body.
An example of the “Discrete Values” filter is show above. All the distinct applications in the workload form the set of discrete values to pick from.
Filtering on read operations is a good example for the “Sliding Range Controls” filter. Other common application cases would be “Duration” or “CPU usage”.
The right-hand side is dedicated to the grid that holds the results of the filtered and aggregated data of the analysis. Among the standard features that you can expect from the application, is the ability to expand and collapse each group and/or flexibly sorted the result grid by clicking on the column headers. Or reordering the column by dragging and dropping them in the place where you want them to be.
The lower part of the right-hand result pane is dedicated to either show a detailed explanation of the resource consumption of the batch template that is selected in the upper right grid.
Or show the actual SQL statements that comprise the batch.