Algorithms, Performance

Scheduling Data-Intensive Bags of Tasks in P2P Grids leads to transfers of large input data files, which cause delays in completion times. We propose to combine several existing technologies and patterns to perform efficient data-aware scheduling: (1) use of the BitTorrent P2P file sharing proto-col to transfer data, (2) data caching on computational Re-sources, (3) use of a data-aware Resource selection schedul-ing algorithm similar to Storage Affinity, (4) a new Task selection scheduling algorithm (Temporal Tasks Grouping), based on the temporally grouped scheduling of Tasks shar-ing input data files. Data replication is also discussed. The proposed approach does not need an overlay network or Predictive Communications Ordering, making our opera-tional implementation of a P2P Grid middleware easily de-ployable in unstructured P2P networks. Experiments show that performance gains are achieved by combining BitTor-rent, caching, Storage Affinity and Temporal Tasks Group-ing. This work can be summarized as combining P2P Grid computing and P2P data transfer technologies