Name
Abstract | ||
|---|---|---|
Horizontal scaling in cloud systems provides adaptation in the virtual infrastructure of the services according to the changing loads. By automatic scaling the system reacts based on measured metrics regarding the operational properties of the virtual infrastructure, however, it is not easy to decide when to initiate the scaling. This paper evaluates CPU utilization based automatic scaling and proposes a new method where direct feedback from the clients is incorporated into the decision when a scaling operation has to be started. We demonstrate the usability of this new method in a Video on Demand service case study. We show that using client feedback on the perceived playback quality supports more accurate decision making when to scale, avoiding unnecessarily scale out events that also leads to cost savings. |
Year | Venue | Keywords |
|---|---|---|
2018 | 2018 41ST INTERNATIONAL CONFERENCE ON TELECOMMUNICATIONS AND SIGNAL PROCESSING (TSP) | cloud, scaling, feedback, OpenStack |
Field | DocType | Citations |
On demand,Cloud systems,CPU time,Computer science,Usability,Real-time computing,Direct feedback,Scaling,Scalability,Distributed computing,Cloud computing | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Maliosz, Markosz | 1 | 5 | 4.27 |
| Csaba Simon | 2 | 4 | 4.18 |
| David Balla | 3 | 5 | 2.43 |
| Anh Tra Phan Ngo | 4 | 0 | 0.34 |
| daniel gehberger | 5 | 2 | 2.16 |