Methods of load balancing using primary and stand-by addresses and related load balancers and servers

• US 9,137,165 B2
• Filed: 06/17/2013
• Issued: 09/15/2015
• Est. Priority Date: 06/17/2013
• Status: Active Grant

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: Anand, Prashant, Arisoylu, Mustafa, Ganesh, Jayasenan Sundara, Sawant, Nandan Mahadeo
• Primary Examiner(s): CRUTCHFIELD, CHRISTOPHER M
• Application Number:
  US13/919,411
• Publication Number:
  US 20140369204A1
• Time in Patent Office: 820 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: H04L 47/125 by balancing the load, e.g. traffic engineering
• CPC Class Codes: H04L 47/21 using leaky-bucket
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1004 Server selection for load balancing
• CPC Class Codes: H04L 67/1023 based on a hash applied to IP addresses or costs

Abstract

A first data packet of a data flow may be addressed to a primary address and include information for the data flow and a bucket ID may be computed based on the information. Responsive to the bucket ID mapping to first and second servers and the first data packet being addressed to the primary address, the first data packet may be transmitted to the first server. A second data packet may be received addressed to a stand-by address and including the information for the data flow, and a bucket ID may be computed based on the information with the bucket IDs for the first and second packets being the same. Responsive to the bucket ID for the second data packet mapping to first and second servers and the second data packet being addressed to the stand-by address, the second data packet may be transmitted to the second server.

Claims

1. A method of forwarding data packets from a load balancer to a plurality of servers, the method comprising:

receiving a first data packet of a data flow wherein the first data packet is addressed to a primary address for the load balancer and wherein the first data packet includes information for the data flow;

computing a bucket ID for the first data packet as a function of the information for the data flow;

responsive to the bucket ID for the first data packet mapping to first and second servers of the plurality of servers and responsive to the first data packet being addressed to the primary address for the load balancer, transmitting the first data packet to the first server;

receiving a second data packet of the data flow wherein the second data packet is addressed to a stand-by address for the load balancer and wherein the second data packet includes the information for the data flow;

computing a bucket ID for the second data packet as a function of the information for the data flow wherein the bucket IDs for the first and second packets are the same bucket ID;

responsive to the bucket ID for the second data packet mapping to first and second servers of the plurality of servers and responsive to the second data packet being addressed to the stand-by address for the load balancer, transmitting the second data packet to the second server.

2. The method of claim 1 wherein the data flow is a first data flow and wherein the bucket ID for the first and second data packets is a first bucket ID, the method further comprising:

receiving a third data packet of a second data flow wherein the third data packet is addressed to a primary address for the load balancer and wherein the third data packet includes information for the second data flow;

computing a second bucket ID for the third data packet as a function of the information for the second data flow; and

responsive to the second bucket ID for the third data packet mapping only to a third server of the plurality of servers, transmitting the third data packet to the third server.

3. The method of claim 2 further comprising:

before receiving the first, second, and third data packets, defining a mapping table including a plurality of bucket identifications (IDs) identifying a respective plurality of buckets, wherein the mapping table maps the first bucket ID to the first server as a current server for the first bucket ID, wherein the mapping table maps the first bucket ID to the second server as a new server for the first bucket ID, and wherein the mapping table maps the second bucket ID to only the third server as a current server for the second bucket ID.

4. The method of claim 3 further comprising:

after transmitting the first, second, and third data packets, revising the mapping table so that the mapping table maps the first bucket ID only to the second server as a current server for the first bucket ID, and wherein the mapping table maps the second bucket ID to only the third server as a current server for the second bucket ID.

5. The method of claim 4 further comprising:

after revising the mapping table, receiving a fourth data packet wherein the fourth data packet is addressed to the primary address for the load balancer and wherein the fourth data packet includes information for the first data flow;

computing the first bucket ID for the fourth data packet as a function of the information for the first data flow; and

responsive to the first bucket ID for the fourth data packet mapping only to the second server of the plurality of servers, transmitting the fourth data packet to the second server.

6. The method of claim 1 wherein the primary address comprises a primary Internet Protocol (IP) address for the load balancer, and wherein the stand-by address comprises a stand-by Internet Protocol (IP) address for the load balancer.

7. The method of claim 1 wherein computing the bucket ID for the first data packet comprises performing a hash function on the information for the data flow.

8. The method of claim 7 wherein the information for the data flow includes a data flow identification (ID) for the data flow, and wherein performing the hash function comprises performing the hash function on the data flow ID.

9. A load balancer configured to forward data packets to a plurality of servers, the load balancer comprising:

a network interface configured to receive data packets from an outside network;

a server interface configured to forward data packets to the servers;

a processor coupled to the network interface and the server interface, the processor being configured to,

receive a first data packet of a data flow through the network interface wherein the first data packet is addressed to a primary address for the load balancer and wherein the first data packet includes information for the data flow,

compute a bucket ID for the first data packet as a function of the information for the data flow,

transmitting the first data packet through the server interface to the first server responsive to the bucket ID for the first data packet mapping to first and second servers of the plurality of servers and responsive to the first data packet being addressed to the primary address for the load balancer,

receive a second data packet of the data flow through the network interface wherein the second data packet is addressed to a stand-by address for the load balancer and wherein the second data packet includes the information for the data flow,

compute a bucket ID for the second data packet as a function of the information for the data flow wherein the bucket IDs for the first and second packets are the same bucket ID, and

transmit the second data packet through the server interface to the second server

responsive to the bucket ID for the second data packet mapping to first and second servers of the plurality of servers and responsive to the second data packet being addressed to the stand-by address for the load balancer.

10. The load balancer of claim 9 wherein the data flow is a first data flow and wherein the bucket ID for the first and second data packets is a first bucket ID, and wherein the processor is further configured to,

receive a third data packet through the network interface wherein the third data packet is addressed to the primary address for the load balancer and wherein the third data packet includes information for a second data flow,

compute a second bucket ID for the third data packet as a function of the information for the second data flow, and

transmit the third data packet to a third server responsive to the second bucket ID for the third data packet mapping only to the third server of the plurality of servers.

11. The load balancer of claim 10 further comprising;

memory coupled to the processor, wherein before receiving the first, second, and third data packets, the memory is configured to define a mapping table including a plurality of bucket identifications (IDs) identifying a respective plurality of buckets, wherein the mapping table maps the first bucket ID to the first server as a current server for the first bucket ID, wherein the mapping table maps the first bucket ID to the second server as a new server for the first bucket ID, and wherein the mapping table maps the second bucket ID to only the third server as a current server for the second bucket ID.

12. A method of receiving data from a load balancer at a server, the method comprising:

defining a server flow table for the server, the server flow table including data flow identifications for data flows being processed by the server;

receiving a data packet of a data flow at the server from the load balancer wherein the data packet includes information for the data flow;

responsive to a data flow identification of the data flow being omitted from server flow table, computing a bucket identification (ID) for the data packet as a function of the information for the data flow;

responsive to the bucket ID for the data packet mapping to only the first server, adding a data flow identification for the data flow to the server flow table; and

responsive to the bucket ID for the data packet mapping to only the first server, processing the data packet.

13. The method of claim 12 wherein the data packet comprises a first data packet, and wherein the data flow comprises a first data flow, the method further comprising:

after adding the data flow identification for the first data flow to the server flow table, receiving a second data packet of a second data flow at the server from the load balancer wherein the second data packet includes information for the second data flow;

responsive to a second data flow identification of the data flow being omitted from server flow table, computing a bucket identification (ID) for the second data packet as a function of the information for the second data flow, wherein the first and second data flows are different and wherein the bucket identifications for the first and second packets are the same;

responsive to the bucket ID for the first and second data packets mapping to the first server and to a second server and responsive to the bucket ID being mapped to the first server before being mapped to the second server, transmitting a redirect request associated with a stand-by address; and

responsive to the bucket ID for the first and second data packets mapping to the first and second servers and responsive to the bucket ID being mapped to the first server before being mapped to the second server, dropping the second data packet.

14. The method of claim 13 wherein the stand-by address comprises a stand-by Internet Protocol (IP) address and wherein transmitting the redirect request comprises transmitting a Hypertext Transfer Protocol (HTTP) redirect request including the stand-by IP address.

15. The method of claim 13 wherein the second data packet originated from a client device, and wherein transmitting the redirect request comprises transmitting the redirect request associated with the stand-by address to the client device.

16. The method of claim 12 wherein the data packet comprises a first data packet, and wherein the data flow comprises a first data flow, the method further comprising:

receiving a second data packet of a second data flow at the server from the load balancer wherein the second data packet includes information for the second data flow;

responsive to a second data flow identification of the data flow being omitted from server flow table, computing a bucket identification (ID) for the second data packet as a function of the information for the second data flow, wherein the first and second data flows are different and wherein the bucket identifications for the first and second packets are the same;

responsive to the bucket ID for the second data packet mapping to the first server and to a second server and responsive to the bucket ID being mapped to the second server before being mapped to the first server, adding a data flow identification for the second data flow to the server flow table; and

responsive to the bucket ID for the second data packet mapping to the first server and to a second server and responsive to the bucket ID being mapped to the second server before being mapped to the first server, processing the second data packet.

17. The method of claim 12 wherein the data packet is a first data packet, and wherein the data flow is a first data flow, the method further comprising:

receiving a second data packet of a second data flow at the first server from the load balancer wherein the second data packet includes information for the second data flow; and

responsive to a data flow identification of the second data flow being included in the list of data flow identifications for data flows being processed by the first server, processing the second data packet at the first server.

18. The method of claim 12 wherein computing the bucket ID for the data packet comprises performing a hash function on the information for the data flow.

19. The method of claim 18 wherein the information for the data flow includes the data flow identification (ID) for the data flow, and wherein performing the hash function comprises performing the hash function on the data flow ID.

20. A server configured to process data packets, the server comprising:

a load balancer interface configured to received data packets from a load balancer;

a memory configured to store a server flow table for the server, the server flow table including data flow identifications for data flows being processed by the server; and

a processor coupled to the load balancer interface and to the memory, wherein the processor is configured to,

receive a data packet of a data flow through the load balancer interface from the load balancer wherein the data packet includes information for the data flow,

compute a bucket identification (ID) for the data packet as a function of the information for the data flow, responsive to a data flow identification of the data flow being omitted from server flow table,

add a data flow identification for the data flow to the server flow table responsive to the bucket ID for the data packet mapping to only the first server, and

process the data packet responsive to the bucket ID for the data packet mapping to only the first server.

21. The server of claim 20 wherein the data packet comprises a first data packet, wherein the data flow comprises a first data flow, and wherein the processor is further configured to,

receive a second data packet of a second data flow through the load balancer interface from the load balancer wherein the second data packet includes information for the second data flow after adding the data flow identification for the first data flow to the server flow table,

compute a bucket identification (ID) for the second data packet as a function of the information for the second data flow responsive to a second data flow identification of the data flow being omitted from server flow table, wherein the first and second data flows are different and wherein the bucket identifications for the first and second packets are the same,

transmit a redirect request associated with a stand-by address responsive to the bucket ID for the first and second data packets mapping to the first server and to a second server and responsive to the bucket ID being mapped to the first server before being mapped to the second server, and

drop the second data packet responsive to the bucket ID for the first and second data packets mapping to the first and second servers and responsive to the bucket ID being mapped to the first server before being mapped to the second server.