11 January,2017 by Jack Vamvas
When managing large SQL Server Inventories , it’s important for me to measure the level of batch requests on a SQL Server. Knowing this figure I can track changes over time. It will give you a measurable method to track impact of changes .
It will also give you statistical feedback , which is used to make decisions about these servers.
Some typical DBA decisions could be:
2) Should I distribute some of the load onto another server , particuarly if it’s possible a few inefficient queries could severely impact server performance . More on SQL Server scaling techniques
3) Will Senior DBAs be required to apply extra time performance tuning and other extra maintenance tasks to keep the server running efficiently?
4) What are the peak periods?
A solid way of measuring is using Perfmon : SQLServer:SQL Statistics - Batch Requests/sec.
The point of capturing these statistics over time is to create a baseline. There is no universal number which gives you absolute kowledge.
That way you can identify trends over time. If you’re capturing these details every day – you’ll identify highs & lows.
It’s important to capture a number of figures, representing different measurables. Having an efficient method of capturing the Perfmon : SQLServer:SQL Statistics - Batch Requests/sec is key to capturing & storing the data. To give you some ideas : Window Perfmon scripting, SQL Server perfmon and how to perfmon