Microsoft’s SQL Customer Advisory Team (SQLCAT) has a new web site that features Top Ten Lists, Technical Notes and Technical Case Studies about various complex performance, scalability, availability subjects concerning SQL Server 2005 and 2008.

One recent Technical Note that caught my eye was Database Mirroring Log Compression in SQL Server 2008 Improves Throughput by Sanjay Mishra. In a pattern that I seen over and over in my career, compressing data to save network bandwidth is usually a net performance win. You give up some CPU to compress and decompress the data, but the compressed data is sent more quickly over the network (especially on lower bandwidth connections), resulting in higher throughput and better performance.

Sanjay’s Technical Note has a wealth of benchmarks and graphs that illustrate the point pretty well. This past week at the SQL PASS Community Summit, the SQL CAT team was there in force, wearing short sleeved orange shirts as they fielded questions from attendees all week and gave several well attended presentations.