Method for allocating cloud service to servers of data center

• US 9,641,451 B2
• Filed: 07/16/2014
• Issued: 05/02/2017
• Est. Priority Date: 01/23/2014
• Status: Active Grant

• Current Assignee: Acer, Inc.
• Original Assignee: Acer, Inc.
• Inventors: Tsai, Lin-Jiun, Liao, Wan-Jiun
• Primary Examiner(s): Harrell, Robert B
• Application Number:
  US14/332,407
• Publication Number:
  US 20150207755A1
• Time in Patent Office: 1,021 Days
• Field of Search: 709226
• CPC Class Codes: H04L 47/76 using dynamic resource allocation, e.g. in-call renegotiation requested by the user or requested by the network in response to changing network conditions

Abstract

A dynamic planning method for server resources of a data center is provided. The data center includes multiple pods and each pod includes multiple racks configurable with multiple servers. In the method, a rack having a remaining space larger than or equal to d is searched from the pods and used for configuring d servers. If the rack cannot be configured with the d servers, a single pod reallocation is executed on one of the pods to empty a rack so that the remaining space of the rack is larger than or equal to d and configurable with the d servers. If there is no rack that can be emptied, a cross-pod reallocation is executed to empty corresponding server positions of corresponding racks in the pods so that the remaining space of the corresponding server positions is larger than or equal to d and configurable with the d servers.

Claims

1. A dynamic planning method for server resources of a data center, adapted to a data center for allocating a service, wherein the service requests to configure d servers, the data center comprising a plurality of pods, each of the pods comprising a plurality of racks respectively connected to a plurality of edge switches, each of the racks being disposed with a plurality of servers, and the servers are sequentially coupled to a plurality of aggregation switches by the connected edge switches, wherein d is a positive integer, the dynamic planning method comprising:

searching, by the data center, for a rack having a remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack;

if the rack cannot be found, executing, by the data center, a single pod reallocation on one of the pods to empty one of the racks of the pod, in order to facilitate the rack so the remaining resource space of the rack is larger than or equal to d, and allocating, by the data center, the d servers to the rack, wherein the step of executing the single pod reallocation comprises;

building a placement list comprising a plurality of reallocation placements for the pod, the reallocation placements comprising the exchanges of two server positions in the pod, and the exchanges of the corresponding server positions between the pods;

for each of the racks in the pod, representing all of the reallocation placements of the server positions in the rack by using a bipartite graph, and selecting a plurality of non-overlapping reallocation placement sets by using a maximum cardinality bipartite matching algorithm; and

selecting the rack from the reallocation placement sets in the pod having a placement quantity larger than or equal to d, and executing d reallocation placements on the rack, so as to empty the rack; and

if no rack in the pod can be emptied, executing, by the data center, a cross-pod reallocation on all of the pods to empty the corresponding server positions of the corresponding racks in the pods, in order for the remaining resource space of the corresponding server positions to be larger than or equal to d, and allocating, by the data center, the d servers to the corresponding server positions.

2. The dynamic planning method according to claim 1, wherein the step of executing the cross-pod reallocation on all of the pods to empty the corresponding server positions of the corresponding racks in the pods, in order to facilitate the remaining resource space of the corresponding server positions to be larger than or equal to d comprises:

building a placement list comprising a plurality of reallocation placements for each of the pods, wherein the reallocation placements comprises every empty server position and the exchanges of two server positions in each of the pods;

calculating a reallocable pod quantity for each of the corresponding server positions (a, e) between the pods, wherein a represents an assigned label of the corresponding aggregation switch, and e represents an assigned label of the corresponding edge switch; and

selecting the server positions (a, e) having a reallocable pod quantity larger than or equal to d, and executing the reallocation placements on the pods having reallocable server positions (a, e), so as to empty the corresponding server positions (a, e) of the corresponding racks in the pods.

3. The dynamic planning method according to claim 2, wherein the step of selecting the server positions (a, e) having the reallocable pod quantity larger than or equal to d comprises:

selecting a first server position (a, e) placed in front of the server positions (a, e) having the reallocable pod quantity larger than or equal to d.

4. The dynamic planning method according to claim 2, wherein the step of executing the reallocation placements on the pods having reallocable server positions (a, e), so as to empty the corresponding server positions (a, e) of the corresponding racks in the pods comprises:

executing d reallocation placements placed ahead on the pods having reallocable server positions (a, e), so as to empty the corresponding server positions (a, e) of the corresponding racks in the pods.

5. The dynamic planning method according to claim 2, wherein the step of building the placement list comprising the reallocation placements for each of the pods comprises:

adding every empty server position in the pod to the placement list;

adding the exchanges of any two server positions in the plurality of corresponding server positions between the racks of the pod to the placement list; and

adding the exchanges of any two server positions not belonging to the same rack in the pod and not belonging to the same corresponding server positions between the racks to the placement list.

6. The dynamic planning method according to claim 1, wherein the step of searching for the rack having the remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack comprises:

searching for at least one pod having a total remaining resource space that is the most from the pods, wherein the total remaining resource space is a sum of the remaining resource spaces of all of the racks in each of the pods; and

selecting a first pod placed in front of the pods to allocate the service.

7. The dynamic planning method according to claim 6, wherein after the step of selecting the first pod placed in front of the pods to allocate the service, the method further comprises:

determining whether the total remaining resource space of the first pod is smaller than d; and

if the total remaining resource space is smaller than d, terminating the allocation of the service.

8. The dynamic planning method according to claim 6, wherein the step of executing the single pod reallocation on one of the pods comprises:

executing the single pod reallocation on the first pod.

9. The dynamic planning method according to claim 1, wherein the step of searching for the rack having the remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack further comprises:

allocating the d servers to the server positions placed in front of the rack.

10. The dynamic planning method according to claim 1, wherein the step of selecting the rack from the reallocation placement sets in the pod having the placement quantity larger than or equal to d comprises:

selecting a first rack placed in front of the racks from the reallocation placement sets in the pod having the placement quantity larger than or equal to d.

11. The dynamic planning method according to claim 1, wherein the step of executing the d reallocation placements on the rack, so as to empty the rack comprises:

executing the d reallocation placements in front of the reallocation placements of the rack, so as to empty the rack.

12. The dynamic planning method according to claim 1, wherein the step of building the placement list comprising the reallocation placements for the pod comprises:

adding the exchanges of any two server positions in the plurality of corresponding server positions between the racks of the pod to the placement list;

adding the exchanges of any two server positions not belonging to the same rack in the pod and not belonging to the same corresponding server positions between the racks to the placement list; and

adding the exchanges of any two corresponding server positions between the pods to the placement list.

13. A dynamic planning method for server resources of a data center, adapted to a data center for allocating a service, wherein the service requests to configure d servers, the data center comprising a plurality of pods, each of the pods comprising a plurality of racks respectively connected to a plurality of edge switches, each of the racks being disposed with a plurality of servers, and the servers are sequentially coupled to a plurality of aggregation switches by the connected edge switches, wherein d is a positive integer, the dynamic planning method comprising:

searching, by the data center, for a rack having a remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack;

if the rack cannot be found, executing, by the data center, a cross-rack reallocation on one of the pods to empty one of the corresponding server positions of the racks in the pod, in order for the remaining resource space of the server position to be larger than or equal to d, and allocating, by the data center, the d servers to the server position of each of the racks, wherein the step of executing the cross-rack reallocation comprises;

building a placement list comprising a plurality of reallocation placements for the pod, wherein the reallocation placements comprises the exchanges of two server positions in the pod, and the exchanges of the corresponding server positions between the pods;

for the corresponding server position in the rack of the pod, representing all of the reallocation placements of the server position by using a bipartite graph, and selecting a plurality of non-overlapping reallocation placement sets by using a maximum cardinality bipartite matching algorithm; and

selecting the server position having a placement quantity of the reallocation placement sets larger than or equal to d from the corresponding server positions of the rack, and executing d reallocation placements on the server position, so as to empty the server position; and

if no server position in the pod can be emptied, executing, by the data center, a cross-pod reallocation on all of the pods to empty the corresponding server positions of the corresponding racks in the pods, in order for the remaining resource space of the corresponding server positions to be larger than or equal to d, and allocating, by the data center, the d servers to the corresponding server positions.

14. The dynamic planning method according to claim 13, wherein the step of searching for the rack having the remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack comprises:

searching for at least one pod having a total remaining resource space that is the most from the pods, wherein the total remaining resource space is a sum of the remaining resource spaces of all of the racks in each of the pods; and

selecting a first pod placed in front of the pods to allocate the service.

15. The dynamic planning method according to claim 14, wherein after the step of selecting the first pod placed in front of the pods to allocate the service, the method further comprises:

determining whether the total remaining resource space of the first pod is smaller than d; and

if the total remaining resource space is smaller than d, terminating the allocation of the service.

16. The dynamic planning method according to claim 14, wherein the step of executing the cross-rack reallocation on one of the pods comprises:

executing the cross-rack reallocation on the first pod.

17. The dynamic planning method according to claim 13, wherein the step of searching for the rack having the remaining resource space larger than or equal to d from the pods, for allocating the d servers to the rack further comprises:

allocating the d servers to the server positions placed in front of the rack.

18. The dynamic planning method according to claim 13, wherein the step of selecting the server position having the placement quantity of the reallocation placement sets larger than or equal to d from the corresponding server positions of the rack comprises:

selecting a first server position placed in front of the server positions having the placement quantity of the reallocation placement sets larger than or equal to d from the corresponding server positions of the rack.

19. The dynamic planning method according to claim 13, wherein the step of executing the d reallocation placements on the server position, so as to empty the server position comprises:

executing the d reallocation placements placed in front of the reallocation placements for the server position, so as to empty the server position.

20. The dynamic planning method according to claim 13, wherein the step of building the placement list comprising the reallocation placements for the pod comprises:

adding the exchanges of any two server positions in each rack of the pod to the arrangement list;

adding the exchanges of any two server positions not belonging to the same rack in the pod and not belonging to the same corresponding server positions between the racks to the arrangement list; and

adding the exchanges of any two corresponding server positions between the pods to the arrangement list.