After becoming a statistician at UC Berkeley, Jeff worked for IBM 36+ years as a z/OS mainframe performance engineer. First as a JES3 developer he helped convert MVS from a 24-bit to 31-bit operating system and then later as a Db2 developer he helped advance it to 64-bit.
While working in DFSMS Development for 15 years, Jeff architected many solutions, including data set level I/O statistics and storage caching solutions. While working in Db2 Development for 16 years, Jeff architected solutions such as List Prefetch Optimizer and Db2 Castout Accelerator to improve the performance of Db2 for z/OS.
Jeff authored numerous papers about the performance of Db2 and the I/O subsystem, and appeared frequently at IDUG, z/OS Expo and z/OS Technical University conferences in North America and Europe.
Resources created by Jeff Berger
Db2 for z/OS Buffer Pool Simulation
Db2 performs best when it has lots of memory. Memory is not free, but large memory can save money by reducing CPU consumption while at the same time reducing Db2 transaction response time.
Understanding Physical Sequential Datasets
Oftentimes it is sufficient to analyze I/O performance without needing to understand the datasets. The SMF 42 records are useful to obtain a deeper understanding of the datasets and the applications that are using the storage.