Just staring at the output from for instance sys.dm_io_virtual_file_stats is usualy not very helpful. Creating delta's from the data and then representing it as a graph helps to understand that data. Over the years I've tried various ways to look at output from SQL Server for performance tuning reasons. In this session I will share a couple of them with you. We will look at virtual_file_stats, waitstats, the buffer cache etc. and we will visualize them in various ways to make it easier to draw conclusions. The visualisation techniques vary from simple graphs that combine throughput and latency to more complicated scenarios that I saw being used by Formula 1 on tv, which translate well to Waitstats and Buffer pool cache usage.
Presented by Andre Kamman at SQLBits XIV
MP4 Video HD 345 MB