Ingress rate limiting

• US 8,842,536 B2
• Filed: 12/23/2010
• Issued: 09/23/2014
• Est. Priority Date: 12/23/2010
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Brocade Communications Systems, Inc. (Broadcom, Inc.)
• Inventors: Gupta, Somesh
• Primary Examiner(s): Yao, Kwang B
• Assistant Examiner(s): NGO, NGUYEN HOANG
• Application Number:
  US12/977,933
• Publication Number:
  US 20120163175A1
• Time in Patent Office: 1,370 Days
• Field of Search: 370/229, 370/230, 370/230.1, 370/231, 370/232, 370/235, 370/236, 370/359, 370/419, 370/463
• US Class Current: 370/235
• CPC Class Codes: H04L 47/10 Flow control; Congestion control

Abstract

A network device monitors the traffic of individual flows through one of its ingress ports and, if the traffic volume exceeds a predetermined threshold, signals for a reduction in data traffic volume transmitted to that ingress port from one or more source devices. Example signals may include without limitation a unicast congestion message sent to the source of a flow, an Explicit Congestion Notification to one or more source devices, and the dropping of packets by the receiving device. In response to such signals, one or more of the source devices decrease the transmission rate of data traffic to the receiving device.

Claims

1. A network device comprising:

a controller for connection in a computer, the computer including a plurality of virtual machines, with each of the plurality of virtual machines having a maximum incoming data rate, the controller configured to detect data traffic received at an ingress port of the controller and directed to each of the plurality of virtual machines exceeding the maximum incoming data rate for each of the plurality of virtual machines, wherein the maximum incoming data rate is less than a data traffic capacity of the ingress port of the controller and the controller is further configured to provide a signal to at least one source of the data traffic directed to a given virtual machine of the plurality of virtual machines to reduce the data output of the at least one source to the virtual machine, responsive to detecting the data traffic received at the ingress port of the controller exceeding the maximum incoming data rate for the given virtual machine.

2. The network device of claim 1, wherein the controller provides the signal to all of the sources of the data traffic directed to the at least one virtual machine.

3. The network device of claim 1, wherein the controller provides the signal to the source of all of the sources of the data traffic directed to the at least one virtual machine that has the highest incoming data rate.

4. The network device of claim 1, wherein the maximum incoming data rate corresponds to a maximum data processing capability of the at least one virtual machine.

5. The network device of claim 1, wherein the maximum incoming data rate corresponds to a service level agreement applying to the at least one virtual machine.

6. The network device of claim 1, wherein the controller is further configured to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine.

7. The network device of claim 6, wherein the signal includes a unicast congestion message.

8. The network device of claim 1,

wherein the controller is further configured to provide a congestion signal to the at least one virtual machine responsive to detecting the data traffic received at the ingress port of the controller exceeding the maximum incoming data rate for the at least one virtual machine and

wherein the at least one virtual machine is configured to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine responsive to receiving the congestion signal from the controller.

9. The network device of claim 8, wherein the signal includes an explicit congestion notification.

10. The network device of claim 1, wherein the network device is a server and the controller is a network interface controller of the server.

11. One or more non-transitory processor-readable storage media encoding processor-executable instructions for executing in a processor a computer process comprising:

detecting data traffic received at an ingress port of a controller and directed to each of a plurality of virtual machines coupled to the controller exceeding a maximum incoming data rate of a given one of the plurality of virtual machines, wherein the maximum incoming data rate is less than a data traffic capacity of the ingress port of the controller; and

signaling at least one source of the data traffic directed to each of the given virtual machines to reduce the data output of the at least one source to each of the given virtual machines, responsive to the detecting operation.

12. The one or more non-transitory processor-readable storage media of claim 11, wherein the controller provides the signal to all of the sources of the data traffic directed to the at least one virtual machine.

13. The one or more non-transitory processor-readable storage media of claim 11, wherein the controller is further configured to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine.

14. The one or more non-transitory processor-readable storage media of claim 11, further comprising:

providing a congestion signal to the at least one virtual machine responsive to detecting the data traffic received at the ingress port of the controller exceeding the maximum incoming data rate for the at least one virtual machine to allow the at least one virtual machine to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine responsive to receiving the congestion signal from the controller.

15. A method comprising:

detecting data traffic received at an ingress port of a controller and directed to each virtual machine of a plurality of virtual machines coupled to the controller exceeding a maximum incoming data rate of a given one of the plurality of virtual machines, wherein the maximum incoming data rate is less than a data traffic capacity of the ingress port of the controller; and

signaling at least one source of the data traffic directed to each of the given virtual machines to reduce data output of the at least one source to each of the given virtual machines, responsive to the detecting operation.

16. The method of claim 15, wherein the controller provides the signal to all of the sources of the data traffic directed to the at least one virtual machine.

17. The method of claim 15, wherein the controller is further configured to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine.

18. The method of claim 15, further comprising:

providing a congestion signal to the at least one virtual machine responsive to detecting the data traffic received at the ingress port of the controller exceeding the maximum incoming data rate for the at least one virtual machine to allow the at least one virtual machine to originate the signal to the at least one source to reduce the data traffic output of the at least one source to the at least one virtual machine responsive to receiving the congestion signal from the controller.