We use the newly-released TPCx-HCI benchmark to characterize the performance and resilience properties of Hyper-Converged Infrastructure clusters. We demonstrate that good performance on an HCI cluster requires delivering all properties of high IOPS, low latencies, low CPU overhead, and uniform access to data from all Nodes. We show that unless the cluster can quickly and efficiently rebalance the VMs after a change in the workload, performance will be severely impacted. We use the data accessibility test of TPCx-HCI to show how performance is impacted by rebuilding traffic after a Node goes down, and how long it takes for the rebuilding to finish.
CITATION STYLE
Taheri, H. R., Little, G., Desai, B., Bond, A., Johnson, D., & Kopczynski, G. (2019). Characterizing the performance and resilience of HCI clusters with the TPCx-HCI benchmark. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11135 LNCS, pp. 58–70). Springer Verlag. https://doi.org/10.1007/978-3-030-11404-6_5
Mendeley helps you to discover research relevant for your work.