A Fine-Grained SDN Rule Table Partitioning and Distribution
Yutaro Yoshikawa, Masayuki Arai · 2019
In Software-Defined Networking (SDN) it is important to efficiently partition the rule table into sub-tables and distribute them to the switches over the network in order to reduce storage size of each switch. In this paper we propose a fine-grained rule table partitioning, where the original rule table is partitioned into n sub-tables and every switch has k sub-tables. We formulate the problem as a satisfiability (SAT) problem, and use SAT solver to obtain an optimal sub-table allocation that minimize the storage size under a given k. Results indicate that the proposed scheme with k = 2 can reduce the storage size required to switches in comparison to the conventional scheme, that corresponds to k = 1.