A Component Assignment Framework for Improved Capacity and Assured Performance in Web Portals

Abstract

Abstract. Web portals hosting large-scale internet applications have become popular due to the variety of services they provide to their users. These portals are developed using component technologies. Important design challenges for developers of web portals involve (1) determining the component placement that maximizes the number of users/requests (capacity) without increasing hardware resources and (2) maintaining the performance within certain bounds given by service level agreements (SLAs). The multitude of behavioral patterns presented by users makes it hard to identify the incoming workloads. This paper makes three contributions to the design and evaluation of web portals that address these design challenges. First it introduces an algorithmic framework that combines bin-packing and modeling-based queuing theory to place components onto hardware nodes. This capability is realized by the Component Assignment Framework for multi-tiered internet applications (CAFe). Second, it develops a component-aware queuing model to predict web portal performance. Third, it provides extensive experimental evaluation using the Rice University Bidding System (RUBiS). The results indicate that CAFe can identify opportunities to increase web portal capacity by 25% for a constant amount of hardware resources and typical web application and user workloads.

A capacity planning process for performance assurance of component-based distributed systems

“…The service demands are thus computed by profiling each single component on the different machines by running with a single client in the system. In previous work [14], it was seen that CPU activity of RUBiS increases with load compared to memory and bandwidth which remain fairly constant. Thus CPU is mainly considered here.…”

“…To highlight the need for load balancing, we reproduce Figure 10 from our earlier work [14] in which processor utilizations of two servers in RUBiS are shown. In the machine DB_SRV, only one component called SearchItemsByCat takes up 70% of processor time when the number of clients reaches around 1,300.…”

“…Resource allocation algorithms developed in prior research [6,7,14] one of the components could then be placed onto BT_SRV so that the overall earlier load of 70+40 = 110 can be balanced across the two servers (55 each). Such allocation makes it possible to handle more clients since now the utilization of both servers can be increased to around 70.…”