Adaptive WebView Materialization

Abstract

Dynamic content generation poses huge resource demands on web servers, creating a scalability problem. WebView Materialization, where web pages are cached and constantly refreshed in the background, has been shown to ameliorate the scalability problem without sacrificing data freshness. In this work we present an adaptive online algorithm to select which WebViews to materialize, that realizes the trade-off between Quality of Service and Quality of Data. Our algorithm performs very close to the static, off-line optimal algorithm, and, under rapid workload changes, it outperforms the optimal.