Optimizing allocation of on-demand resources using performance zones

1. A method comprising:

receiving a user request for allocation of an on-demand resource within a data center having a plurality of performance zones each logically connected by a logical network, each performance zone formed of and corresponding to a logical collection of computing devices, wherein communication between computing devices within a performance zone is optimized as compared to communications between performance zones;

determining whether the user request includes a performance zone request for a performance zone with which the user is associated and if so, determining whether a resource corresponding to the requested on-demand resource is available within the requested performance zone;

assigning to the user the resource within the requested performance zone if the resource is available, and otherwise causing the user request to fail, and if the user request does not include a performance zone request, assigning to the user a second resource corresponding to the requested on-demand resource within a second performance zone outside of the performance zone if the resource is not available within the performance zone;

wherein resources of the user in the performance zone and the second on-demand resource in the second performance zone appear to the user as a single logical network; and

providing an equal level of service to communications between the second on-demand resource and the on-demand resource of the user within the performance zone and to communications between the on-demand resource of the user within the performance zone and a third on-demand resource of the user within the performance zone.

2. The method of claim 1 , wherein communication between computing devices within a performance zone is secondarily optimized with respect to communications with computing devices in a particular geographic location outside the performance zone.

3. The method of claim 1 , further comprising providing a tunnel with a predetermined bandwidth between the second resource and at least one resource of the user within the performance zone.

4. The method of claim 1 , further comprising provisioning the on-demand resource and sending information regarding the provisioned on-demand resource to the user.

5. The method of claim 1 , further comprising provisioning the on-demand resource in the performance zone according to a scheduler having a plurality of criteria, wherein a performance zone criteria is weighted greater than other criteria of the scheduler.

6. The method of claim 1 , wherein the on-demand resource comprises a storage resource, and wherein the storage resource is within the performance zone with which the user is associated.

7. A non-transitory computer readable medium comprising instructions that when executed enable a system to perform a method comprising:

receiving a user request for allocation of an on-demand resource within a data center having a plurality of performance zones each logically connected by a logical network, each performance zone formed of and corresponding to a logical collection of computing devices, wherein communication between computing devices within a performance zone is optimized as compared to communications between performance zones;

determining whether the user request includes a performance zone request for a performance zone with which the user is associated and if so, determining whether a resource corresponding to the requested on-demand resource is available within the requested performance zone; assigning to the user the resource within the requested performance zone if the resource is available, and otherwise causing the user request to fail; and if the user request does not include a performance zone request, assigning to the user a second resource corresponding to the requested on-demand resource within a second performance zone outside of the performance zone if the resource is not available within the performance zone, and providing a tunnel with a predetermined bandwidth between the second resource and at least one resource of the user within the performance zone;

wherein resources of the user in the performance zone and the second on-demand resource in the second performance zone appear to the user as a single logical network, and wherein the method further comprises providing an equal level of service to communications between the second on-demand resource and the on-demand resource of the user within the performance zone and to communications between the on-demand resource of the user within the performance zone and a third on-demand resource of the user within the performance zone.

8. The non-transitory computer readable medium of claim 7 , wherein communication between computing devices within a performance zone is secondarily optimized with respect to communications with computing devices in a particular geographic location outside the performance zone.

9. The non-transitory computer readable medium of claim 7 , further comprising instructions that when executed enable the system to perform a method further comprising provisioning the on-demand resource and sending information regarding the provisioned on-demand resource to the user.

10. The non-transitory computer readable medium of claim 7 , further comprising instructions that when executed enable the system to perform a method further comprising provisioning the on-demand resource in the performance zone according to a scheduler having a plurality of criteria, wherein a performance zone criteria is weighted greater than other criteria of the scheduler.

11. A multi-tenant data center comprising:

a plurality of hosts each corresponding to a server having a processor, memory, network interface circuitry and mass storage, wherein groups of the plurality of hosts are segmented into a plurality of server collections, each of the plurality of server collections coupled to a switch of an aggregation layer, and wherein switches of the aggregation layer are in turn coupled to switches of a core layer, wherein a customer of the multi-tenant data center is provided a logical performance zone in which a plurality of on-demand resources of the multi-tenant data center allocated to the customer appear to the customer as a single logical network having a given quality of service (QoS), wherein communication between hosts within a server collection designated as a logical performance zone is optimized as compared to communications between hosts within the logical performance zone and hosts outside the logical performance zone, and at least some of the on-demand resources allocated to the customer are located in different ones of the plurality of server collections and are coupled to other on-demand resources located in different server collections via dedicated tunnels having guaranteed bandwidth performance via an auto-creation of a network reservation having the guaranteed bandwidth;

a scheduler to provision the on-demand resources allocated to the customer within the plurality of server collections, wherein the scheduler is to provision the on-demand resources allocated to the customer according to a plurality of criteria, wherein a performance zone criteria is weighted highest among the plurality of criteria; and

a logical switch to enable mappings between the on-demand resources of the single logical network, the single logical network configured to span across at least some of the plurality of server collections, wherein the on-demand resources allocated to the customer each include a virtual switch to provide a direct connection to the virtual switch of others of the on-demand resources allocated to the customer within a physical performance zone.

12. The multi-tenant data center of claim 11 , wherein the logical switch is allocated to the customer to provide interconnection between virtual switches of the single logical network.

13. The multi-tenant data center of claim 11 , wherein the scheduler is to weight a request by the customer for allocation of a first on-demand resource within the logical performance zone with a first value and to weight a second allocation criteria with a second value.

14. The multi-tenant data center of claim 13 , wherein the scheduler is to schedule the first on-demand resource to the logical performance zone based on the weighted allocation request and the weighted second allocation criteria.