|
…
|
||
|---|---|---|
| .. | ||
| SqlPlugin | ||
| SqlPluginWithProcessingPipeline | ||
| README.md | ||
| SampleSqlTraceFile.xml | ||
| SqlPlugin.sln | ||
README.md
Introduction
This solution is the source code for the SQL plugin created during a demonstration on how to use the SDK.
Step-by-step video walkthroughs on creating this plugin can be found at:
- Part 1: creating a simple SQL custom data source and table
- Part 2: recreating the SQL table using the data-processing pipeline
This plugin consists of two custom data sources (each defined in its own project/assembly/namespace). Both
custom data sources implement logic for parsing .xml files that contain SQL Server Profiler
trace data (i.e. a .xml file created by exporting a .trc file).
Each project contains one table that displays information contained within the .xml file parsed.
The simple SQL custom data source's table displays all events inside the XML (except
the Trace Start and Trace End events). The data-processing pipeline table receives
its data by querying a data cooker that receives only SQL:BatchStarting and SQL:BatchCompleted
events from the source parser. As such, it only displays those events. The data cooker this
table depends on is also responsible for mapping raw SqlEvent data classes to SqlEventWithRelativeTimestamp
data classes, which augment the raw data with a Timestamp representing when a given event
occurred relative to the Trace Start event. For the simple custom data source, this
conversion happens inside the ProcessAsyncCore method of the custom data processor.
Both custom data sources only look for/display a subset of fields which a SQL trace may contain. This was done to make the plugin easier to understand and to better serve as a live demonstration. A full implementation of this plugin would likely have a different structure.