An Efficient Algorithm of Context-Clustered Microservice Discovery
Huan Liu, Zhiying Cao, Xiuguo Zhang · Proceedings of the 2nd International Conference on Computer Science and Application Engineering · 2018
Using context1 is helpful to find services that satisfy users' requirements more accurately. In this paper context is classified into service context and user context and an algorithm of context-clustered microservice discovery (ACCMD) is proposed. Firstly, microservices are clustered according to the similarity of service context and a candidate service set is initialized by matching the request with service clusters. Then, according to the similarity of user context, users whoever used and are using the candidate services are clustered and the candidate service set is refined by matching the requester's information with the context of users in these clusters. Finally, Quality of Service (QoS) and requester's preference are used to filter candidate services and the results are inverted indexed and returned to the requester. Comparative experiments show that the ACCMD can increase user suitability of discovery and help filter out microservices quickly.