Merging Topic Groups of a Publish/Subscribe System in Causal Order
Yuta Yamamoto, Naohiro Hayashibara · 2017
Message ordering is one of the essential problems in distributed systems and applications. In spite of the importance, most of publish/subscribe systems do not guarantee the ordering of messages and events because of its cost which may give an impact on the overall performance. When we suppose dynamic systems, such as peer-to-peer systems, the implementation of the message ordering would be pretty difficult and costly. In this paper, we propose an efficient algorithm to merge separated topic groups with preserving causal order in publish/subscribe systems on unstructured peer-to-peer systems. We also evaluate the proposed algorithm in terms of the cost for merging in our experiments.