Title :
Graph-Based Caching for Server-Push Enabled Adaptive Hypertext Applications
Author :
Daniel Thommes;Ansgar Gerlicher;Qi Wang;Christos Grecos
Author_Institution :
Responsive Media Experience Group, Stuttgart Media Univ., Stuttgart, Germany
Abstract :
The inherent structure of applications defined with hypertext markup languages is that of a graph. Push-enabled protocols like HTTP/2 drive the need for solutions allowing the server to traverse this graph and preemptively push related resources to the client. In addition to this, user interface adaptation to various mobile devices belongs to the server´s tasks. Taking into account the diversities in client hardware, software and user preferences we designed a solution to reuse once adapted resources for similar configurations. In this paper we propose a new caching architecture for these purposes that is based on the application of graph database systems. With different approaches we show how to increase their performance when used as resource caches. In our evaluation the proposed implementation reduced the amount of adapted resources by 89%. In a real-world scenario it achieved twice the throughput of a traditional implementation and used 40% less heap memory.
Keywords :
"Servers","Time factors","Protocols","Optimization","Layout","User interfaces","Hardware"
Conference_Titel :
Computational Science and Engineering (CSE), 2015 IEEE 18th International Conference on
DOI :
10.1109/CSE.2015.22