Name
Abstract | ||
|---|---|---|
DC/OS is a widely used distributed operating system that abstracts the resources of light-weighted virtualized datacenters, which is based on Mesos distributed systems kernel and user space services such as Marathon. It automates resource management and process scheduling, thus significantly impacts the performance of datacenters. In this paper, we propose Themis, a flexible, automatic and distributed framework, to evaluate the performance and scalability across both DC/OS and virtualization layer. We can integrate most emerging scale-out datacenter workloads into this framework, and get an easily understandable score that scales with underlying system capacity using a configurable controlled strategy. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1109/CLUSTER.2016.78 | 2016 IEEE International Conference on Cluster Computing (CLUSTER) |
Keywords | Field | DocType |
Themis,scalable performance evaluation framework,DC/OS,distributed operating system,light-weighted virtualized datacenters,Mesos distributed systems kernel,user space services,Marathon,resource management,process scheduling,distributed framework,virtualization layer,underlying system capacity,configurable controlled strategy | Virtualization,Distributed operating system,Computer science,Scheduling (computing),Server,Real-time computing,User space,Distributed computing,Resource management,Parallel computing,Operating system,Cloud computing,Scalability | Conference |
ISSN | ISBN | Citations |
1552-5244 | 978-1-5090-3654-7 | 0 |
PageRank | References | Authors |
0.34 | 2 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhengmin Li | 1 | 0 | 0.68 |
| Di Zhang | 2 | 4 | 3.21 |
| Xinran Liu | 3 | 0 | 0.34 |
| Bin Sun | 4 | 1 | 0.69 |
| Zhicheng Yao | 5 | 4 | 1.13 |
| Xiufeng Sui | 6 | 0 | 0.34 |