Parallel Distributed Infrastructure for Minimization of Energy

ParaDIME presents at IEEE Cloud

Sat, 2013-06-29

ParaDIME member Thomas Knauth from TU Dresden presented work done in context of the the ParaDIME project at the IEEE Cloud conference in Santa Clara, CA, USA. The work deals with ways to improve the replication performance for strongly consistent, geo-replicated data stores. The topic is relvant to ParaDIME because we aim to distribute computations across multiple data centers. However, distribution across data centers creates additional challenges for the timely and consistent distribution of data updates. Improving the data store's performance will also benefit the energy efficency. Increasing the number of data store operations per time slice also increases the time that  storage components can be put into low-power modes.