Nowadays, with the growth of heterogeneity of network equipmentand software, and involved a variety of advanced network technology, thecomplexity of computer network have rapidly increased. Especially demandsfor high-performance network service in advanced data-intensive scientific re-search are dramatically increased. In order to cope with the network perfor-mance issue, this paper proposes the end-to-end (ETE) network performancemanagement framework based on Case-based Reasoning with the case libraryand multi-agent integrated with perfSONAR as well as large-scale networkflow monitoring. It provides a sophisticated framework for both network op-erator and user to systemically identify ETE network performance issues thatare detected, diagnosed and recovered. To validate the framework, the retrievaleffectiveness is demonstrated with modeling and implementing the casebase inthe real experimental environment, a national research network of Korea.