Monday, October 02, 2006

Grid traces

Apparently, our recent study ( A. Iosup, C. Dumitrescu, D.H.J. Epema, H. Li, L. Wolters, How are Real Grids Used? The Analysis of Four Grid Traces and Its Implications, The 7th IEEE/ACM International Conference on Grid Computing (Grid), Barcelona, September 28-29, 2006 ) was well received by the Grid computing research community. One of the most important outcomes of this situation is that our work towards the Grid Workloads Archive (GWA) has made a nice and decided step forward.

As much of the research in grids today is made with simulations based on empirical workloads, the GWA is much needed. This is because years of theoretical and practical experience show that studies based on empirical workloads have little or even no relevance for the actual users -- a large-scale distributed system may behave much better than another under different workload assumptions (see the 1990s dispute on this between Eager, Liny, and Harchol-Balter; the two latter arguably have shown that preiously disconsidered migration mechanisms are actually extremely valuable for realistic workloads, while they behave poorly for specially constructed (empirical) workloads).

To conclude, mark the day of 28.Sep.2006 as the first community step towards the GWA, and implicictly towards rigurous workload preparation.


Post a Comment

<< Home