Performance Management for 1,000 Linux servers on System z

This presentation is an extension to the previous session on capacity planning. Performance analysis for 1,000 Linux servers is not done via "top", it requires the ability to easily manage performance problems in any subsystem at the hypervisor layer.  This presentation will show where performance data comes from in a z/VM environment, and show case studies of how installations with 1,000 virtualized Linux servers manage their performance.

Tracks: Linux Systems Management and z/VM Systems Programming