Experiments‎ > ‎

KOPFire: elasticity experiments with KOPI on BonFIRE

The basis of the KOPFire experiment is the KOPI Plagiarism Search Portal of MTA SZTAKI DSD. KOPI is a nationwide plagiarism service in Hungary. KOPI works asynchronously: it accepts requests in the form of uploaded documents, which is checked for copied content over various databases. After this, a report is sent to the user containing the copied parts and their original sources. KOPI service is implemented as a multi-layer infrastructure which means an interconnected set of virtual hosts with different capabilities. KOPI typically suffers from bursts of incoming requests, and therefore would benefit from automatic, elastic scaling of its service infrastructure. 

The experiment will investigate the costs and effects of various scaling operations and their possible combinations in an elasticity strategy. The final goal of the elasticity strategy is to stabilize end-to-end QoS (the quality of experience) around desired values for the processing of incoming plagiarism search requests while keeping resource usage at a minimum. The ideal strategy has to effectively detect and avoid the bottlenecks in the system, by duplicating, stopping or moving the appropriate components, and furthermore it also has to detect capacity overhead and downscale the infrastructure accordingly.



Added-value of the BonFIRE infrastructure

During the experiments we need large storage capacities and hosts with various processing power. Furthermore, elasticity will be tested through cloud federations. A homogeneous way for monitoring the complex scenario and collecting measured data is also essential for us. The simultaneous properties of large amount of resources, a federated cloud environment, ubiquitous monitoring and elasticity facilities altogether are almost impossible to get anywhere else. 

Impact on the BonFIRE project

The problems investigated in this experiment are quite common for a class of on-line services. Some characteristics of this service class include asynchronous response (because the minimum time to respond is too long), and federations with similar services in order to join efforts. Typically, such services have local or national specifics disabling the emergence of large multi-national solutions. 
We think it is important to see how BonFIRE can be applied for such a real-life service, and also to see how the service behaves in the advanced environment of BonFIRE. The results about elastic queue processing can be useful for many other scenarios as well which use large monolithic software components and execute complex time-consuming calculations.