What I have experienced as a Microsoft Data Platform TSP in the past several weeks since the 2011 PASS Summit is that most of the discussion, testing, excitement and energy around the new SQL Server version (now SQL Server 2012, previously SQL Server Code-Named Denali) has centered around the big-ticket items:
- Always On
- Power View (formerly Project Crescent)
- Column-based Indexes
- SQL Appliances
- SQL Server Developer Tools (Visual Studio Integration formerly Project Juneau)
- Pricing Changes
I’ve written in this blog (you can search this from the WordPress search bar at top) about other “2nd-tier” updates to the SQL Server platform that will ship with the GA of SQL Server 2012 in early 2012 such as Sequence Objects.
I want to introduce you to another nice new feature in SQL Server that is something that I have heard customers ask about for a while now: Distributed Replay. This is the ability to capture workload from multiple SQL Servers from an application perspectives and be able to replay those for development & testing purposes. It makes sense that this is a feature that took some time to make its way natively into the product. In the past, I have seen some very clever and effective solutions that customers have wired together using downloaded 3rd party tools, open source generators and replaying SQL Profiler traces.
Those work well enough. But now you have a tool that can capture workload from more than 1 SQL Server at a time and to have a tool that can play those transactions back with different load targets. This will be very helpful for regression testing, load testing and staging among other areas.
Jonathan Kehayias at SQL Skills put up a blog on how to get in installed and configured (there is a server & a client component) and I like James Serra’s write up as an overview to learn a little bit more.