SYSTEMS AND METHODS FOR DISTRIBUTED HASH TABLE IN A MULTI-CORE SYSTEM

• US 20100322250A1
• Filed: 06/22/2009
• Published: 12/23/2010
• Est. Priority Date: 06/22/2009
• Status: Active Grant

• Current Assignee: Citrix Systems, Inc. (Cloud Software Group)
• Original Assignee: Citrix Systems, Inc. (Cloud Software Group)
• Inventors: Khemani, Prakash, Shetty, Anil, Suganthi, Josephine
• Granted Patent: US 8,737,407 B2
• Time in Patent Office: Days
• US Class Current: 370/395.32
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: H04L 41/50 Network service management, e.g. ensuring proper service fulfilment according to agreements

Abstract

The present invention is directed towards systems and methods for using a distributed hash table to maintain the same configuration and resource persistency across a plurality of cores in a multi-core system. The distributed hash table includes a plurality of partitions, each partition being owned by a respective core of the multi-core system. A core may establish resources in the partition it owns. A core may request other cores to establish resources in the partitions they own and send resource information to the core. The core may locally cache the resource information.

Claims

1. A method of maintaining resource information via a distributed table for resources established by a plurality of packet engines, each packet engine executing on a respective core in a multi-core system, the method comprising:

a) establishing, by a multi-core system, a table comprising a plurality of partitions, each partition of the plurality of partitions assigned to a respective core of a plurality of cores of the multi-core device;

b) storing, by a first packet engine of a first core of the plurality of cores, resource information for resources established by the first packet engine to a first partition of the plurality of partitions, and storing, by a second packet engine of a second core of the plurality of cores, resource information established by the second core to a second partition of the plurality of partitions;

c) receiving, by the first packet engine of the first core, a request identifying a first resource; and

d) determining, by the first packet engine of first core, that the first resource is stored in a partition other than the first partition.

2. The method of claim 1, wherein step (a) further comprises dividing, by the multi-core device, the partition table into the plurality of partitions based on a number of cores of the multi-core device.

3. The method of claim 1, wherein step (c) further comprises receiving, by the first packet engine, the request on a virtual server that is configured to provide resource persistency.

4. The method of claim 1, wherein step (c) further comprises generating, by the first packet engine, a hash of transport layer connection information of the request.

5. The method of claim 4, wherein step (d) further comprises determining, by the first packet engine, that the first resource is not stored in the first partition based on the hash.

6. The method of claim 4, wherein step (d) further comprises identifying, by the first packet engine, that the first resource was established by the second core based on the hash.

7. The method of claim 1, wherein step (d) further comprises accessing, by the first packet engine, the first partition of the partition table to determine whether the first resource is stored in the first partition.

8. The method of claim 1, further comprising receiving, by the by the first packet engine, a second request identifying a second resource, determining that the second resource is stored in the first partition and a virtual server on the first core has a status of up, and responsive to the determination, providing the second request to the virtual server, wherein the first resource is a first session and the second resource is a second session.

9. The method of claim 1, further comprising receiving, by the first packet engine, a second request identifying a second resource, determining that the second resource should be stored in the first partition and does not exist in the first partition, and responsive to the determination, creating the second resource.

10. The method of claim 1, further comprising sending, by the first core, a message to the second core, the message requesting information to create the first resource on the first core.

11. The method of claim 10, further comprising determining, by the second packet engine, that the first resource exists in the second partition of the partition table, identifying a status of a service selected by the second packet engine for the first resource is up, and sending a response to the first packet engine identifying the service.

12. The method of claim 10, further comprising determining, by the second packet engine, that the first resource does not exist in the second partition of the partition table, and responsive to the determination creates the first resource, selects a service for the first resource and sends a response to the first core identifying the service.

13. The method of claim 10, further comprising receiving, by the first core, a response to the message from the second core, determining a state of the service on the first core is up, and caching a copy of the first resource in the first partition.

14. A system of maintaining resource information via a distributed partition table for resources established by a plurality of packet engines, each packet engine executing on a respective core in a multi-core system, the system comprising:

a partition table established by a multi-core system, the partition table comprising a plurality of partitions, each partition being assigned to a respective core of the multi-core system;

a first packet engine executing on a first core of the multi-core system that stores information for resources established by the first packet engine to a first partition in the partition table, receives a request identifying a first resource, determines that the first resource is not stored in the first partition, and identifies a second partition that stores the first resource; and

a second packet engine of a second core of the multi-core system that stores information for resources established by the second packet engine to the second partition in the partition table.

15. The system of claim 14, wherein the partition table is divided into the plurality of partitions based on a number of cores of the multi-core system.

16. The system of claim 14, wherein the first packet engine receives the request on a virtual server configured to provide resource persistency.

17. The system of claim 14, wherein the first packet engine generates a hash of transport layer connection information of the request.

18. The system of claim 17, wherein the first packet engine determines that the first resource is not stored in the first partition based on the hash.

19. The system of claim 17, wherein the first packet engine determines that the first resource was established by the second core based on the hash.

20. The system of claim 14, wherein the first packet engine accesses the first partition to determine whether the first resource is stored in the first partition.

21. The system of claim 14, wherein the first packet engine receives a second request identifying a second resource, determines that the second resource is stored in the first partition and a virtual server on the first core has a status of up, and provides the second request to the virtual server in response to the determination.

22. The system of claim 14, wherein the first packet engine receives a second request identifying a second resource, determines that the second resource does not exist in the first partition and the second resource should be stored in the first partition, and creates the second resource in response to the determination.

23. The system of claim 14, wherein the first core sends a message to the second core, the message requesting information to create the first resource on the first core.

24. The system of claim 23, wherein the second packet engine determines that the first resource exists in the second partition, identifies that a status of a service associated with the first resource is up, and sends a response to the first core identifying the service.

25. The system of claim 23, wherein the second packet engine determines that the first resource does not exist in the second partition, creates the first resource in response to the determination, selects a service for the first resource, and sends a response to the first core identifying the service.

26. The system of claim 23, wherein the first packet engine receives a response to the message from the second core, determines that a state of a service associated with the first resource is up, and caches a copy of the first resource in the first partition.