Content based load balancer

• US 8,554,790 B2
• Filed: 12/18/2007
• Issued: 10/08/2013
• Est. Priority Date: 12/18/2007
• Status: Active Grant

• Current Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Original Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Inventors: Turk, Mladen
• Primary Examiner(s): Richardson, James E
• Application Number:
  US11/959,349
• Publication Number:
  US 20090157678A1
• Time in Patent Office: 2,121 Days
• Field of Search: 707/780, 707/782, 707/728, 709/219, 709/238
• US Class Current: 707/780
• CPC Class Codes: G06F 16/21 Design, administration or maintenance of databases
• CPC Class Codes: G06F 16/24552 Database cache management

Abstract

A content based load balancing system receives a request for data provided by a resource. The content based load balancing system searches a content history cache for a content history cache entry corresponding to the requested data. The content based history cache then selects a resource node to service the request based on the content history cache entry corresponding to the data.

Claims

1. A computer-implemented method comprising:

receiving a request for data by a computing device executing a load balancing module, wherein the data is provided by a plurality of resource nodes;

searching, by the computing device, a content history cache for one or more content history cache entries corresponding to the data, at least one of the one or more content history cache entries comprising an indicator of a resource node of the plurality of resource nodes that serviced an associated previous request, wherein the content history cache is not located on the plurality of resource nodes;

responsive to finding the one or more content history cache entries, selecting a first resource node of the plurality of resource nodes to service the request in accordance with a first load balancing algorithm, the first load balancing algorithm being a content based load balancing algorithm, wherein the selecting comprises;

responsive to finding a single content history cache entry, selecting the first resource node associated with the single content history cache entry; and

responsive to finding a plurality of content history cache entries, selecting the first resource node based on a weighting value and a validity value of at least one of the found plurality of matching content history cache entries;

responsive to failing to find the one or more content history cache entries, selecting a second resource node of the plurality of resource nodes to service the request in accordance with a second load balancing algorithm; and

forwarding the request to the selected first resource node or the selected second resource node.

2. The computer-implemented method of claim 1, wherein the request comprises an identifier of the data, the identifier being a uniform resource locator (URL).

3. The computer-implemented method of claim 1, wherein finding the one or more content history cache entries corresponding to the data comprises determining a nearest non-exact match between an identifier of the data from the request and an identifier in at least one of the one or more content history cache entries, the non-exact match comprising a pattern match.

4. The computer-implemented method of claim 1, further comprising:

updating at least one of the one or more content history cache entries with response data from the first resource node.

5. The computer-implemented method of claim 4, wherein the updating comprises:

modifying at least one of the weighting value and the validity value of the at least one content history cache entry.

6. The computer-implemented method of claim 1, wherein the second load balancing algorithm is a non-content based load balancing algorithm.

7. The computer-implemented method of claim 1, further comprising:

modifying a validity of at least one of the one or more content history cache entries in response to predetermined criteria.

8. A non-transitory computer readable medium, having instructions stored therein, which when executed, cause a computing device to perform a set of operations comprising:

receiving a request for data by a load balancing module running on the computing device, wherein the data is provided by a plurality of resource nodes;

searching, by the load balancing module, a content history cache for one or more content history cache entries corresponding to the data, at least one of the one or more content history cache entries comprising an indicator of a resource node of the plurality of resource nodes that serviced an associated previous request, wherein the content history cache is not located on the plurality of resource nodes;

responsive to finding the one or more content history cache entries, selecting a first resource node to service the request in accordance with a first load balancing algorithm, the first load balancing algorithm being a content based load balancing algorithm, wherein the selecting comprises;

responsive to finding a single content history cache entry, selecting the first resource node associated with the single content history cache entry; and

responsive to finding a plurality of content history cache entries, selecting the first resource node based on a weighting value and a validity value of at least one of the found plurality of matching content history cache entries;

responsive to failing to find the one or more content history cache entries, selecting a second resource node of the plurality of resource nodes to service the request in accordance with a second load balancing algorithm; and

forwarding the request to the selected first resource node or the selected second resource node.

9. The non-transitory computer readable medium of claim 8, wherein the request comprises an identifier of the data, the identifier being a uniform resource locator (URL).

10. The non-transitory computer readable medium of claim 8, wherein finding the one or more content history cache entries corresponding to the data comprises determining a nearest non-exact match between an identifier of the data from the request and an identifier in at least one of the one or more content history cache entries, the non-exact match comprising a pattern match.

11. The non-transitory computer readable medium of claim 8, having further instructions therein, which when executed, cause the computer to perform a further set of operations, further comprising:

updating at least one of the one or more content history cache entries with response data from the first resource node.

12. The non-transitory computer readable medium of claim 11, wherein the updating comprises:

modifying the weighting value and the validity value of the at least one content history cache entry.

13. The non-transitory computer readable medium of claim 8, wherein the second load balancing algorithm is a non-content based load balancing algorithm.

14. The non-transitory computer readable medium of claim 8, having further instructions stored therein, which when executed perform a set of operations, further comprising:

modifying the validity value of at least one of the one or more content history cache entries in response to predetermined criteria.

15. A system comprising:

a memory device; and

a processor coupled to the memory device, wherein the processor is configured to;

receive a request for data, wherein the data is provided by a plurality of resource nodes;

search a content history cache for one or more content history cache entries corresponding to the data, at least one of the one or more content history cache entries comprising an indicator of a resource node of the plurality of resource nodes that serviced an associated previous request, wherein the content history cache is not located on the plurality of resource nodes;

responsive to finding the one or more content history cache entries, select a first resource node to service the request in accordance with a first load balancing algorithm, the first load balancing algorithm being a content based load balancing algorithm, wherein the selecting comprises;

responsive to finding a single content history cache entry, selecting the first resource node associated with the single content history cache entry; and

responsive to finding a plurality of content history cache entries, selecting the first resource node based on a weighting value and a validity value of at least one of the found plurality of matching content history cache entries;

responsive to failing to find the one or more content history cache entries, select a second resource node of the plurality of resource nodes to service the request in accordance with a second load balancing algorithm; and

forward the request to the selected first resource node or the selected second resource node.

16. The system of claim 15, wherein the processor is further configured to parse the incoming resource request to determine a requested data identifier.

17. The system of claim 15, wherein the processor is further configured to update the validity value and the weight value of at least one of the one or more content history cache entries in response to receiving a resource node response to the incoming request.

18. The system of claim 15, wherein the processor is further configured to periodically update the validity value of at least one of the one or more content history cache entries in response to predetermined criteria.