Network　virtualization　has　recently　emerged　as　a　promising　solution　for　diversifying　the　future　Internet　architecture　into　separate　virtual　networks　(VNs).　The　problem　of　efficiently　embedding　multiple　independent　VNs　over　a　common　substrate　infrastructure　is　a　challenging　problem　on　cloud　computing　platforms　and　large-scale　future　network　testbeds.　To　solve　this　problem,we　take　advantage　of　aMap-Reduce　framework,　decomposing　a　VN　request　into　meta-requests,　using　a　link-priority　algorithm　and　pheromone　transmission　based　on　multi　intelligent　route　nodes,　which　have　the　ability　to　distribute　meta-requests　and　collect　meta-request　embedding　results　in　the　substrate,　and　assignVNs　to　the　substrate　physical　network　in　a　distributed　and　parallelmanner.Adistributed　and　parallelVN　embedding　protocol　is　proposed　to　communicate　and　exchange　messages　among　substrate　nodes　to　achieve　successful　embedding.　Results　of　implementation　and　a　performance　evaluation　of　the　distributed　and　parallel　VN　embedding　algorithm　in　terms　of　embedding　time,　acceptance　ratio,　produced　messages　and　revenue　and　cost　are　presented　at　the　end　of　this　paper.　©　2013　Trade　Science　Inc.　-　INDIA.