Scaling computing resources in a cluster

• US 9,898,347 B1
• Filed: 03/15/2017
• Issued: 02/20/2018
• Est. Priority Date: 03/15/2017
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP SE
• Inventors: Gupta, Vipul
• Primary Examiner(s): Kim, Dong
• Application Number:
  US15/459,473
• Time in Patent Office: 342 Days
• Field of Search: None
• CPC Class Codes: G06F 9/4881 Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
• CPC Class Codes: G06F 9/5005 to service a request
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5044 considering hardware capabilities
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: H04L 67/1004 Server selection for load balancing
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/53 using third party service providers

Abstract

Systems and methods are provided for receiving a request for an allocation of a task in a cluster comprising a plurality of client nodes, determining a node type for the task, based on mapping the task to a list of categories, wherein at least two of categories in the list of categories overlap in range. The systems and methods further providing for searching for available client nodes based on the node type for the task to select a client node to allocate the task, determining a zone of the selected client node, wherein the zone is mapped to the list of categories, determining a wait algorithm associated with the zone of the selected node, and contacting the selected client node and passing the task and the wait algorithm to the selected client node.

Claims

1. A computerized method, comprising:

receiving, at a manager server node for a cluster comprising a plurality of client nodes, a request for an allocation of a task;

determining, by at least one processor of the manager server node, a node type for the task, based on mapping the task to a list of categories, wherein values of at least two of categories in the list of categories overlap in range;

searching, by the manager server node, for available client nodes based on the node type for the task to select a client node to allocate the task;

determining, by the at least one processor of the manager server node, a zone of the selected client node, wherein the zone is mapped to the list of categories;

determining, by the at least one processor of the manager server node, a wait algorithm associated with the zone of the selected node; and

contacting, by the manager server node, the selected client node and passing the task and the wait algorithm to the selected client node.

2. The method of claim 1, wherein the categories in the list of categories comprise at least a central processing unit (CPU) usage category and an input/output (I/O) usage category.

3. The method of claim 2, wherein the CPU usage category comprises at least one subcategory from a list comprising:

idle, average, and overloaded.

4. The method of claim 2, wherein the CPU usage category comprises at least two subcategories with values that overlap in range.

5. The method of claim 2, wherein the I/O usage category comprises at least one subcategory from a list comprising:

negligible, light, medium, and heavy.

6. The method of claim 1, wherein determining a node type for the task, based on mapping the task to a list of categories, further comprises determining a central processing unit (CPU) usage category and an input/output (I/O) usage category for the task.

7. The method of claim 1, wherein determining the wait algorithm for the task comprises determining that a wait algorithm should be computed based on a status of the selected client node.

8. The method of claim 1, further comprising:

updating allocation details with the task assigned to the selected client node.

9. The method of claim 1, further comprising:

receiving a status message for the selected client node;

determining that a zone associated with the selected client node is different from a stored zone for the client node;

updating records for the cluster with the different zone;

determining a wait algorithm for the selected client node based on the different zone; and

sending the wait algorithm to the selected client node.

10. The method of claim 1, wherein searching for available client nodes based on the node type for the task to select a client node to allocate the task comprises determining which available client nodes have a status associated with the node type for the task.

11. The method of claim 1, wherein the client node processes the task using the wait algorithm.

12. A server computer comprising:

at least one processor; and

a computer-readable medium coupled with the at least one processor, the computer-readable medium comprising instructions stored thereon that are executable by the at least one processor to cause the server computer to perform operations comprising;

receiving a request for an allocation of a task in a cluster comprising a plurality of client nodes;

determining a node type for the task, based on mapping the task to a list of categories, wherein values of at least two of categories in the list of categories overlap in range;

searching for available client nodes based on the node type for the task to select a client node to allocate the task;

determining a zone of the selected client node, wherein the zone is mapped to the list of categories;

determining a wait algorithm associated with the zone of the selected node; and

contacting the selected client node and passing the task and the wait algorithm to the selected client node.

13. The server computer of claim 12, wherein the categories in the list of categories comprise at least a central processing unit (CPU) usage category and an input/output (I/O) usage category.

14. The server computer of claim 13, wherein the CPU usage category comprises at least one subcategory from a list comprising:

idle, average, and overloaded; and

wherein CPU usage category comprises at least two subcategories with values that overlap in range.

15. The server computer of claim 13, wherein the I/O usage category comprises at least one subcategory from a list comprising:

negligible, light, medium, and heavy.

16. The server computer of claim 12, wherein determining a node type for the task, based on mapping the task to a list of categories, further comprises determining a central processing unit (CPU) usage category and an input/output (I/O) usage category for the task.

17. The server computer of claim 12, wherein determining the wait algorithm for the task comprises determining that a wait algorithm should be computed based on a status of the selected client node.

18. The server computer of claim 12, the operations further comprising:

receiving a status message for the selected client node;

determining that a zone associated with the selected client node is different from a stored zone for the client node;

updating records for the cluster with the different zone;

determining a wait algorithm for the selected client node based on the different zone; and

sending the wait algorithm to the selected client node.

19. The server computer of claim 12, wherein searching for available client nodes based on the node type for the task to select a client node to allocate the task comprises determining which available client nodes have a status associated with the node type for the task.

20. A non-transitory computer-readable medium comprising instructions stored thereon that are executable by at least one processor to cause a computing device to perform operations comprising:

receiving a request for an allocation of a task in a cluster comprising a plurality of client nodes;

determining a node type for the task, based on mapping the task to a list of categories, wherein values of at least two of categories in the list of categories overlap in range;

searching for available client nodes based on the node type for the task to select a client node to allocate the task;

determining a zone of the selected client node, wherein the zone is mapped to the list of categories;

determining a wait algorithm associated with the zone of the selected node; and

contacting the selected client node and passing the task and the wait algorithm to the selected client node.