Snooze - Open-source, Scalable, Autonomic, and Energy-efficient VM Management for Private Clouds

Snooze is an open-source, scalable, autonomic, and energy-efficient  virtual machine (VM) management framework for private clouds. Similarly  to other VM management frameworks such as Nimbus, OpenNebula,  Eucalyptus, and OpenStack it allows to build compute infrastructures  from virtualized resources. Particularly, once installed and configured  users can submit and control the life-cycle of a large number of VMs.  However, contrary to existing frameworks for scalability and fault  tolerance, Snooze employs a self-organizing and healing (based on Apache ZooKeeper) hierarchical  architecture. Moreover, it performs distributed VM management and is  designed to be energy efficient. Therefore, it implements features to  monitor and estimate VM resource (CPU, memory, network Rx, network Tx)  demands, detect and resolve overload/underload situations, perform  dynamic VM consolidation through live migration, and finally power  management to save energy. Last but not least, it integrates a generic  scheduler which allows to implement any VM placement algorithms. The  system can be either used to manage production data centers or as an  experimental testbed for advanced (i.e. requiring live migration  support) VM placement algorithms.

Snooze is one of the core results of Eugen Feller`s PhD thesis under the  supervision of Dr. Christine Morin at the INRIA MYRIADS project-team. The prototype  has been verified on the Grid`5000 as well as on the EDF (Électricité de  France) experimental testbeds. Results will be presented on well known international conferences IEEE CLOUD and CCGrid 2012. The design  and development was funded by the French Agence Nationale de la Recherce  (ANR) project EcoGrappe under the contract number ANR-08-SEGI-008-02.