Optimization for multi-thread data-flow software
EPEW'11 Proceedings of the 8th European conference on Computer Performance Engineering
| Hi-index | 0.00 |
This work presents a special class of a data flow oriented optimization tool that finds the optimal number of threads for multi-thread software. Threads are assumed to encapsulate concurrent executable key functionalities, are connected through finite capacity queues, and require certain hardware resources. We show how a combination of measurement and calculation, based on Queueing Theory, leads to an algorithm which recursively determines the best combination of threads, i.e. the best configuration of the multi-thread software on a specific host. The algorithm proceeds on the directed graph of a queueing network which models this software. Optimization towards hardware consolidation, where CPU cores, memory, disk space and speed, and network bandwidth are constraints, but also towards throughput is described. Two experiments on different SUN machines verify our optimization approach.