Popular open source frameworks are used today to automat ically combat attack threats in DCCNs. However, they cannot visualize logical conditions well enough in representing network protocol vulnera bilities in attacks such as ARP poisoning, DNS spoofing, and SYN flood ing, and they do not support advanced types of communication, such as wireless, vehicular, and tethering communications on networks. To address the above shortcomings, we present an AND/OR attack graph based security model for DCCNs, which is capable of modeling multi ple attack types (e.g., ARP poisoning, spoofing, man-in-the-middle, and many others). A novel efficient algorithm is proposed to extract the most vulnerable attack path from the AND/OR graph. Unlike previous heuris tic pathfinding algorithms, the proposed algorithm runs in polynomial time and efficiently handles positive-length cycles. We demonstrate the effectiveness of the proposed model and the new algorithm on several testbed problems implementing a network architecture of IT and indus trial components.