Session awareness issues for next generation cluster based network load balancing frameworks

Abstract

While a lot of researches focused on how to efficiently spread the offered network load on the available cluster resources, less interest has been granted to the impact of the used mechanisms on the reliable execution of the upper layer services. On the other hand, emerging NGN services as well as some of the already familiar services involve multiple flows during the lifespan of a single endto-end session, hence, raising the challenge of session awareness while processing the incoming network traffic.In this paper, we grasp the need for fine grained session awareness to efficiently allocate the cluster resources to the offered network traffic. The analysis of load balancing scenarios of some representative IP services provides us with solid reasons to use deep packet inspection to achieve fine grained network traffic load distribution, and to meet NAT and firewall traversal constraints as well.

SARA: A session aware infrastructure for high performance next generation cluster-based servers

“…Datagrams which are subject to content inspection are those being exchanged over the signalling flows. Indeed, in [3] we showed that multiple-flow based sessions such as video streaming sessions negotiate data flow identifiers and control data flow identifiers using messages exchanged over the main signalling flow. A searched pattern is a particular application level protocol header field holding the information necessary to identify the possible next incoming flows associated to an already established session.…”

“…The cluster global load is considered as critical when it goes beyond the third threshold 3 T . Rather than interrupting the already established sessions leading them to be restarted from scratch, we suggest to instantaneously degrade their QoS.…”

“…In practice, the rejection of the incoming traffic is triggered each time interval i T computed as a function of the dropping probability p as shown in (3).…”

“…Previous works on clusterbased network traffic load balancing [1,2] provide exclusively flow aware solutions which do not guarantee the building of robust and scalable IP service frameworks. Conversely, most of the next generation IP services are built upon a session model which involves multiple and heterogeneous flows required for the signalling and for the data exchange all along the session lifespan [3]. Typical examples include media streaming and voice over IP.…”