Name
Abstract | ||
|---|---|---|
Software streaming is a form of on-demand software distribution services, which means a program need not be installed on the client for execution but parts of it can be delivered as needed. This paper has analyzed a collection of real traces of desktop applications from a production deployment, and located some opportunities for caching data: First, several local cache strategies have been compared and adjusted to fit the access pattern of data. Second, a special cooperative file caching mechanism is designed between clients and server(s). In detail, the cooperative caching mechanism employs the application-level semantics to trace cache status for each application rather than each file, which remarkably reduces the management overheads while the hint accuracy is still high. In addition, we present DES3, the DistributEd Storage for Software Streaming. It adopts the above optimizations; tests show that DES3 can greatly reduce the server load and give acceptable client latency. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1007/978-3-319-03859-9_40 | ICA3PP |
Keywords | Field | DocType |
application virtualization,distributed storage,software as services | Software deployment,Cache,Latency (engineering),Computer science,Computer network,Software,Application virtualization,Distributed computing,Parallel computing,Distributed data store,Smart Cache,Operating system,Software distribution | Conference |
Volume | Issue | Citations |
8285 LNCS | PART 1 | 1 |
PageRank | References | Authors |
0.38 | 9 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Youhui Zhang | 1 | 202 | 28.36 |
| Peng Qu | 2 | 13 | 2.68 |
| Yanhua Li | 3 | 12 | 2.68 |
| Hongwei Wang | 4 | 11 | 8.68 |
| Weimin Zheng | 5 | 1889 | 182.48 |