Method for resource optimized network virtualization overlay transport in virtualized data center environments
View Patent ↗A method for resource optimized network virtualization overlay transport in a virtualized data center environment includes an optimized virtualized transport mechanism based on MPLS-TP tunneling technology. The transport mechanism may include a usage monitoring and resource usage advertisement mechanism based on IGP-TE protocol extension. Also, the transport mechanism may include a path computation engine (PCE) based optimal path computation for a virtualized transport tunnel used to carry VM-to-VM traffic across the virtualized data center. Additionally, the transport mechanism may include a SDN controller for provisioning and control of virtualized machines and a virtualized transport system.
1. A method of determining an optimal placement of a virtual machine in a virtual data center network, wherein the method is implemented by a Path Computation Element (PCE), the method comprising:
monitoring resource utilization by a plurality of virtual machines;
receiving a virtual machine utilization summary from a server;
updating a resource database of virtual machine utilization within the virtualized data center network based on the received virtual machine utilization summary;
receiving a request associated with a first virtual machine; and
determining and executing an optimal placement of the first virtual machine within the virtual data center network based on the updated resource database, wherein the determining the optimal placement comprises determining a data center in the virtual data center and a server in the data center that satisfies policy and resource constraints of the first virtual machine, and wherein the optimal placement is determined from a plurality of physical data centers and a plurality of physical servers in the plurality of physical data centers,
wherein the resource database is maintained as a graph with vertices comprising servers in the virtual data center network and with edges indicating resource utilization attributes from the received virtual machine utilization summary in the graph, wherein the determining the optimal placement comprises utilizing an optimization applied to the graph to find an optimal vertex which satisfies the policy and resource constraints of the first virtual machine, and wherein the policy and resource constraints comprise a plurality of processing power, I/O bandwidth, memory size requirement, virtual disk space limit, and application performance requirement.
2. The method according to claim 1 , wherein the PCE utilizes algorithmic computation calculations to find the optimal placement of the first virtual machine.
3. The method according to claim 2 , wherein the PCE utilizes algorithmic computation calculations to find an optimal re-routing of an existing virtual machine.
4. The method according to claim 2 , wherein the PCE dynamically maps the resource utilization status of a plurality of servers in the virtual data center network based on a status message received from a virtual data center monitor.
5. The method according to claim 2 , wherein the PCE runs policy constraint combinatorial optimization algorithms to find a free virtual machine slot on a first server that will satisfy the policy constraints.
6. The method according to claim 2 , wherein the virtual data center network is selected from one of a group of networks, the group comprising a Virtual extensible Local Area Network (VXLAN), a Network Virtualization Generic Routing Encapsulation (NV-GRE) network, a media access control (MAC)-in-MAC network and other Layer 2 over a Layer 3 network.
7. The method according to claim 1 , wherein the graph is maintained in a similar manner as a traffic engineering database.
8. The method according to claim 1 , wherein the monitoring comprises listening to traffic engineering advertisements from one or more of OSPF-TE, GMPLS-OSPF-TE, ISIS-TE, and GMPLS-ISIS-TE.
9. The method according to claim 1 , wherein the PCE is configured to monitor network resource advertisements in the virtual data center and to communicate to the server comprising a Hypervisor for the virtual machine utilization summary.
10. An apparatus for determining an optimal placement of a virtual machine in a virtual data center network, the apparatus comprising:
a circuit configured to monitor resource utilization by a plurality of virtual machines;
a circuit configured to receive a virtual machine utilization summary from a server;
a circuit configured to update a resource database of virtual machine utilization within the virtualized data center network based on the received virtual machine utilization summary;
a circuit configured to receive a request associated with a first virtual machine; and
a circuit configured to determine and execute an optimal placement of the first virtual machine within the virtual data center network based on the updated resource database, wherein determining the optimal placement comprises determining a data center in the virtual data center and a server in the data center that satisfies policy and resource constraints of the first virtual machine, and wherein the optimal placement is determined from a plurality of physical data centers and a plurality of physical servers in the plurality of physical data centers,
wherein the resource database is maintained as a graph with vertices comprising servers in the virtual data center network and with edges indicating resource utilization attributes from the received virtual machine utilization summary in the graph, and wherein the determining the optimal placement comprises utilizing an optimization applied to the graph to find an optimal vertex which satisfies the policy and resource constraints of the first virtual machine, and wherein the policy and resource constraints comprise a plurality of processing power, I/O bandwidth, memory size requirement, virtual disk space limit, and application performance requirement.
11. The apparatus according to claim 10 , wherein the graph is maintained in a similar manner as a traffic engineering database.
12. The apparatus according to claim 10 , wherein the monitoring comprises listening to traffic engineering advertisements from one or more of OSPF-TE, GMPLS-OSPF-TE, ISIS-TE, and GMPLS-ISIS-TE.
13. The apparatus according to claim 10 , wherein the apparatus is configured to monitor network resource advertisements in the virtual data center and to communicate to the server comprising a Hypervisor for the virtual machine utilization summary.
14. A Path Computation Engine (PCE) adapted to determine an optimal placement of a virtual machine in a virtual data center network, the PCE comprising:
one or more processors; and
memory comprising instructions that, when executed, cause the one or more processors to monitor resource utilization by a plurality of virtual machines;
receive a virtual machine utilization summary from a server;
update a resource database of virtual machine utilization within the virtualized data center network based on the received virtual machine utilization summary;
receive a request associated with a first virtual machine; and
determine and execute an optimal placement of the first virtual machine within the virtual data center network based on the updated resource database, wherein determining the optimal placement comprises determining a data center in the virtual data center and a server in the data center that satisfies policy and resource constraints of the first virtual machine, and wherein the optimal placement is determined from a plurality of physical data centers and a plurality of physical servers in the plurality of physical data centers,
wherein the resource database is maintained as a graph with vertices comprising servers in the virtual data center network and with edges indicating resource utilization attributes from the received virtual machine utilization summary in the graph, and wherein the determining the optimal placement comprises utilizing an optimization applied to the graph to find an optimal vertex which satisfies the policy and resource constraints of the first virtual machine, and wherein the policy and resource constraints comprise a plurality of processing power, I/O bandwidth, memory size requirement, virtual disk space limit, and application performance requirement.
15. The PCE according to claim 14 , wherein the graph is maintained in a similar manner as a traffic engineering database.
16. The PCE according to claim 14 , wherein the monitoring comprises listening to traffic engineering advertisements from one or more of OSPF-TE, GMPLS-OSPF-TE, ISIS-TE, and GMPLS-ISIS-TE.
17. The PCE according to claim 14 , wherein the PCE is configured to monitor network resource advertisements in the virtual data center and to communicate to the server comprising a Hypervisor for the virtual machine utilization summary.