Seamless Integration of Generic Bulk Operations in Grid Applications

Within grid environments, as the number of remote operations increases, the latencies for these operations become a dominant factor for application performance. Bulk operations represent one potential methodology to reduce latencies for large numbers of similar operations. However, the identification of bulks can pose a non-trivial exercise for application developers, often implying changes in the remote API, and hence direct code modifications to the application. In this work, we identified those properties an API must possess to provide for transparent bulk optimizations, allowing a latency hiding approach which does not require code changes on application level, and without complicating the remote API itself. We have developed a prototype implementation (within the SAGA C++ reference implementation effort), and present performance measurements for bulks of remote file copy operations.