Project: capacity and performance management in storage systems.

Advisors: Eitan Bachmat (CS, BGU), Julian Satran (IBM research, Haifa)

General description: The purpose of the project is to construct a
prototype system that will help customers make storage purchasing and
management decisions based on historical usage data from their own sites.
This is especially important when customers consolidate data from legacy
systems, a major industry trend.

The software will have a component that will read historical metadata
about storage system access patterns from standard industry tools such as
IBM Tivoli, HP openview, EMC control center etc.

The system will contain a very simple algorithmic engine that will use the
data
to assess configurations with Flash, SCSI and SATA drives and choose the
best ones in terms of price, performance and power consumption.  
It will also assessw the benefits of thin provisioning and the potential
energy saving from spin down in remote and locally mirrored systems.

The third component will be a GUI to communicate with the customer.

Requirments: No previous knowledge about storage systems is needed,
what is needed is good intelligence and some code writing and system
design capabilities.
  
Number of persons: The team should consist of 3 students