Title
Dynamic scaling of call-stateful SIP services in the cloud
Abstract
Many cloud technologies available today support dynamically scaling out and back computing services. The predominantly session-oriented nature and the carrier-grade requirements of telco services (such as SIP services) complicate the successful adoption of dynamic scaling in a telco cloud. This paper investigates how to enable dynamic scaling of these telco services in an effective manner, focusing in particular on call-stateful SIP services. First, we present and evaluate two protocols to transparently migrate ongoing sessions between call-stateful SIP servers. These allow to quickly shutdown call-stateful SIP servers in response to a scale back request, removing the need to wait until their ongoing calls have finished. Second, instead of responding to load changes in a reactive manner, this paper explores the potential value of pro-active resource provisioning based on call load forecasting. We propose a self-adaptive Kalman filter to implement short-term call load predictions and combine this with history-based predictions to anticipate future call load changes. We believe that session migration and call load forecasting are two important elements to safely reduce the operational expenditure (OpEx) of a cloudified SIP service.
Year
DOI
Venue
2012
10.1007/978-3-642-30045-5_14
Networking (1)
Keywords
Field
DocType
call load forecasting,future call load change,call-stateful sip service,cloudified sip service,shutdown call-stateful sip server,short-term call load prediction,dynamic scaling,call-stateful sip server,sip service,telco service,elasticity,telecommunication,sip,cloud
Computer science,Server,Dynamic scaling,Computer network,Load forecasting,Provisioning,Kalman filter,Stateful firewall,Operating expense,Cloud computing
Conference
Volume
ISSN
Citations 
7289
0302-9743
5
PageRank 
References 
Authors
0.63
12
4
Name
Order
Citations
PageRank
Nico Janssens1877.75
Xueli An2211.90
Koen Daenen3122.75
Claudio Forlivesi415510.07