Impact of Inter-application Contention in Current and Future HPC Systems
Impact of Inter-application Contention in Current and Future HPC Systems
复制标题
当前和未来 HPC 系统中应用程序间争用的影响
DOI：
--
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
J. Labarta
中科院分区：
文献类型：
--
作者：
Ana Jokanovic;G. Rodríguez;J. Sancho;J. Labarta
胖树网络是当今超级计算机中最流行的间接网络拓扑结构。当前的超级计算机通常在作业调度器的控制下在共享环境中操作，同时执行许多并行应用。这些应用程序之间使用相同网络资源的竞争导致应用程序性能的下降。必须等待网络资源被另一个应用程序的消息占用的应用程序被称为正在经历应用程序间争用。应用程序间的竞争所造成的退化程度是已知的，取决于多个因素：网络拓扑结构，路由方案，任务的位置，等等，请注意，这些因素也是一个直接的应用程序内竞争。我们的工作评估的影响，应用程序间的竞争，实际竞争的HPC工作负载下不同的路由方案在瘦胖树。与以往的作品，主要集中在个别应用程序的性能，我们采取了更系统为中心的观点。我们的工作估计了当前HPC系统中应用程序间争用造成的系统性能损失的数量，我们已经测量到约为10%。我们还提出了在近期和中期的未来HPC系统的应用程序间的竞争的影响，缩放节点的计算能力和网络链路速度可预见的值的预测。我们的研究结果表明，网络速度的提高不需要保持与计算能力的提高一样快的速度，但它仍然需要扩大规模。我们对未来HPC系统的预测表明，即使对于某些应用程序混合，链路速度为40 Gb/s，应用程序间争用也可能导致15%的吞吐量损失。当使用不同的混合运行时，所选应用程序的影响差异会导致以前的作品中描述的性能变化，但我们的工作比注入网络噪声的研究更好地限制了变化。
Fat-tree networks are the most popular topology among indirect networks in today’s supercomputers. Current supercomputers are generally operated in a shared environment under the control of a job scheduler, executing many parallel applications simultaneously. The competition between these applications to use the same network resources causes a degradation in the applications’ performance. The application that has to wait for the network resources occupied by another application’s messages is said to be experiencing inter-application contention. The extent of degradation caused by inter-application contention is known to depend on multiple factors: the network topology, the routing scheme, the task-placement, etc. Note that these factors also a direct intra-application contention. Our work evaluates the impact of inter-application contention for actual competing HPC workloads under different routing schemes in slimmed Fat Trees. In contrast with previous works, which focus mostly on individual application’s performance, we take a more system-centric view. Our work estimates the amount of system performance loss that inter-application contention contributes in current HPC systems, which we have measured to be around a 10%. We also present a projection of the impact of inter-application contention in the near and mid -term future HPC systems, scaling the node computational power and network link speeds to foreseeable values. Our results suggest that the increase in network speed does not need to keep the same fast pace as the increase in computation power, but it still needs to be scaled up. Our projection for future HPC systems shows that interapplication contention can cause a 15% throughput loss even with link speeds of 40 Gb/s for some application mixes. The difference in impact for a chosen application when running with different mixes leads to the performance variability described in previous works, but our work sets a better bound on the variability than studies performed with an injection of network noise.