SYSTEMS AND METHODS FOR LEAST CONNECTION LOAD BALANCING BY MULTI-CORE DEVICE

• US 20150039763A1
• Filed: 07/31/2013
• Published: 02/05/2015
• Est. Priority Date: 07/31/2013
• Status: Active Grant

• Current Assignee: Citrix Systems, Inc. (Cloud Software Group)
• Original Assignee: Citrix Systems, Inc. (Cloud Software Group)
• Inventors: Chaudhary, Aman, Itta, Vishnu, Prakash, Devesh, Muthiah, Manikam, Sridhar, Jaidev
• Granted Patent: US 9,088,501 B2
• Time in Patent Office: Days
• US Class Current: 709/226
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: H04L 47/125 by balancing the load, e.g. traffic engineering
• 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 67/1023 based on a hash applied to IP addresses or costs

Abstract

The present invention is directed towards systems and methods for load balancing by a multi-core device intermediary between clients and services. The device may establish sub-slots in each slot of the device'"'"'s packet engines. The number of sub-slots may correspond to the packet engine count. Each slot may track a different number of active connections allocated to a service. The device may assign a first and second service to each packet engine in a first slot corresponding to no active connections. These services may be assigned to different sub-slots in adjacent packet engines. The device may update, responsive to allocation of a first active connection to the first service, the first service from a sub-slot in the first slot of a first packet engine, to a corresponding sub-slot in a second slot. The second slot may correspond to one active connection allocated to the first service.

Claims

1. A method for load balancing a plurality of connections to a plurality of services across a plurality of packet engines of a multi-core device, the method comprising:

(a) establishing a number of sub-slots in each slot of each of a plurality of packet engines executing on a device intermediary between a plurality of clients and a plurality of services, the number of sub-slots corresponding to a number of the packet engines executing on the device, each of the packet engines executing on a respective core from a plurality of cores on the device, each slot of each packet engine for tracking a different number of active connections allocated to a service accessible via the corresponding packet engine;

(b) assigning, by the device, a first service and a second service each having no active connections, to each of the packet engines in a first slot corresponding to no active connections, the first service and the second service assigned to different sub-slots of the first slot in adjacent packet engines; and

(c) updating, by the device responsive to allocation of a first active connection to the first service, the first service from a sub-slot in the first slot of a first packet engine of the plurality of packet engines, to a corresponding sub-slot in a second slot of the first packet engine, the second slot corresponding to one active connection allocated to the first service.

2. The method of claim 1, comprising updating, by the device responsive to allocation of a first active connection to the second service, the second service from a sub-slot in the first slot of a second packet engine of the plurality of packet engines, to a corresponding sub-slot in a second slot of the second packet engine, the second slot corresponding to one active connection allocated to the second service.

3. The method of claim 1, comprising updating, by the device responsive to removal of the first active connection from the first service, the first service from the sub-slot in the second slot of the first packet engine to the corresponding sub-slot in the first slot of the first packet engine.

4. The method of claim 1, comprising updating, by the device responsive to removal of the first active connection from the first service, the first service from a sub-slot in the Xth slot of the first packet engine to a corresponding sub-slot in the (X−

1)th slot of the first packet engine.

5. The method of claim 4, comprising setting a pointer to a next available sub-slot of the Xth slot if no other service remains in the Xth slot, the pointer for use in allocating a new connection to a service.

6. The method of claim 1, comprising adding, by device, the first service to the first slot of the first packet engine, and if no other existing service has been added prior to the slot, and setting a pointer to a sub-slot of the first slot to which the first service has been added, the pointer for use in allocating a new connection to a service.

7. The method of claim 6, comprising establishing a connections list to track the pointer.

8. The method of claim 1, wherein if the number of services assigned to each packet engine exceeds the number of packet engines, step (b) further comprises assigning a third service to the sub-slot of each packet engine to which the first service is assigned.

9. The method of claim 1, wherein if the number of services assigned to each packet engine is less than the number of packet engines, step (b) further comprises assigning the services to a same group of sub-slots for each packet engine.

10. The method of claim 1, comprising re-assigning one or more remaining services to one or more slots of the packet engines if an existing service is removed or a new service added.

11. A system for load balancing a plurality of connections to a plurality of services across a plurality of packet engines of a multi-core device, comprising:

a plurality of packet engines executing on a device intermediary between a plurality of clients and a plurality of services; and

a load balancer executing on the device, the load balancer;

establishing a number of sub-slots in each slot of each of the plurality of packet engines, the number of sub-slots corresponding to a number of the packet engines executing on the device, each of the packet engines executing on a respective core from a plurality of cores on the device, each slot of each packet engine for tracking a different number of active connections allocated to a service accessible via the corresponding packet engine,

assigning a first service and a second service each having no active connections, to each of the packet engines in a first slot corresponding to no active connections, the first service and the second service assigned to different sub-slots of the first slot in adjacent packet engines, and

updating, responsive to allocation of a first active connection to the first service, the first service from a sub-slot in the first slot of a first packet engine of the plurality of packet engines, to a corresponding sub-slot in a second slot of the first packet engine, the second slot corresponding to one active connection allocated to the first service.

12. The system of claim 11, wherein the load balancer updates, responsive to allocation of a first active connection to the second service, the second service from a sub-slot in the first slot of a second packet engine of the plurality of packet engines, to a corresponding sub-slot in a second slot of the second packet engine, the second slot corresponding to one active connection allocated to the second service.

13. The system of claim 11, wherein the load balancer updates, responsive to removal of the first active connection from the first service, the first service from the sub-slot in the second slot of the first packet engine to the corresponding sub-slot in the first slot of the first packet engine.

14. The system of claim 11, wherein the load balancer updates, responsive to removal of the first active connection from the first service, the first service from a sub-slot in the Xth slot of the first packet engine to a corresponding sub-slot in the (X−

1)th slot of the first packet engine.

15. The system of claim 14, wherein the load balancer sets a pointer to a next available sub-slot of the Xth slot if no other service remains in the Xth slot, the pointer for use in allocating a new connection to a service.

16. The system of claim 11, wherein the load balancer adds the first service to the first slot of the first packet engine, and if no other existing service has been added prior to the slot, and sets a pointer to a sub-slot of the first slot to which the first service has been added, the pointer for use in allocating a new connection to a service.

17. The system of claim 16, wherein the load balancer establishes a connections list to track the pointer.

18. The system of claim 11, wherein if the number of services assigned to each packet engine exceeds the number of packet engines, the load balancer assigns a third service to the sub-slot of each packet engine to which the first service is assigned.

19. The system of claim 11, wherein if the number of services assigned to each packet engine is less than the number of packet engines, the load balancer assigns the services to a same group of sub-slots for each packet engine.

20. The system of claim 11, wherein the load balancer re-assigns one or more remaining services to one or more slots of the packet engines if an existing service is removed or a new service added.