Scalable flow and cogestion control with openflow

Patent metadata

Field	Value
Publication number	US-9503382-B2
Application number	US-201414501945-A
Country	US
Kind code	B2
Filing date	Sep 30, 2014
Priority date	Mar 15, 2013
Publication date	Nov 22, 2016
Grant date	Nov 22, 2016

Abstract

A method of performing congestion management in a network is provided. The method includes monitoring a congestion status at a switch in the network. It is determined at the switch that the congestion status indicates that there is congestion at the switch. Based on the congestion being related to data received at the switch from a source at a first rate, a first message is transmitted from the switch to the source requesting the source to send the data at a second rate that is slower than the first rate. Based on the congestion being related to data requests received at the switch from a destination at third rate, a second message is transmitted from the switch to the destination requesting the destination to request the data at a fourth rate slower than the third rate.

First claim

What is claimed is:

1. A method of performing congestion management in a network, the method comprising:
monitoring, at a physical switch in the network, both data packets and data requests received at the switch over one or more links in the network;
based on the monitoring, determining a congestion status at the switch;
determining, at the switch, that the congestion status indicates that there is congestion at the switch, the determining including examining a plurality of flags in a header field of an entry corresponding to a flow in a flow table at the switch, the plurality of flags including a forward congestion flag that when set indicates that a source of the data packets for the data flow is sending the data packets more quickly than the switch can process the data packets, and a backward congestion flag that when set indicates that a destination of the data packets for the data flow is sending the data requests more quickly than the switch can process the data requests;
based on the forward congestion flag being set and congestion being related to the data packets received at the switch from the source at a first rate, transmitting a first message from the switch to the source, the first message requesting the source to send the data packets at a second rate that is slower than the first rate; and
based on the backward congestion flag being set and congestion being related to the data requests received at the switch from the destination at a third rate, transmitting a second message from the switch to the destination, the second message requesting the destination to send the data requests at a fourth rate that is slower than the third rate.

2. The method of claim 1 , further comprising:
based on transmitting the first message:
monitoring the congestion status at the switch in the network; and
based on determining that the congestion status indicates that there is no longer congestion at the switch related to the data packets received at the switch from the source, transmitting a third message from the switch to the source requesting the source to send the data packets at a fifth rate that is faster than the second rate.

3. The method of claim 1 , further comprising:
based on transmitting the second message:
monitoring the congestion status at the switch in the network; and
based on determining that the congestion status indicates that there is no longer congestion at the switch related to the data requests received at the switch from the destination, transmitting a fourth message to the destination requesting the destination to send the data requests at a sixth rate that is faster than the fourth rate.

4. The method of claim 1 , wherein the switch, the source, and the destination are OpenFlow switches and the network is an OpenFlow network.

5. The method of claim 1 , wherein the switch is an OpenFlow switch.

6. The method of claim 5 , wherein the forward congestion flag is set to indicate congestion at the switch related to the data packets received at the switch based on a queue at the switch that stores the data packets exceeding a specified capacity.

7. The method of claim 5 , wherein the backward congestion flag is set to indicate congestion at the switch related to the data requests received at the switch based on a queue at the switch that stores the data requests exceeding a specified capacity.

8. A computer program product for congestion management in a network, the computer program product comprising:
a tangible storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising:
monitoring, at a physical switch in the network, both data packets and data requests received at the switch over one or more links in the network;
based on the monitoring, determining a congestion status at the switch in the network;
determining, at the switch, that the congestion status indicates that there is congestion at the switch, the determining including examining a plurality of flags in a header field of an entry corresponding to a flow in a flow table at the switch, the plurality of flags including a forward congestion flag that when set indicates that a source of the data packets for the data flow is sending the data packets more quickly than the switch can process the data packets, and a backward congestion flag that when set indicates that a destination of the data packets for the data flow is sending the data requests more quickly than the switch can process the data requests;
based on the forward congestion flag being set and congestion being related to the data packets received at the switch from the source at a first rate, transmitting a first message from the switch to the source, the first message requesting the source to send the data packets at a second rate that is slower than the first rate; and
based on the backward congestion flag being set and congestion being related to the data requests received at the switch from the destination at a third rate, transmitting a second message from the switch to the destination, the second message requesting the destination to send the data requests at a fourth rate that is slower than the third rate.

9. The computer program product of claim 8 , wherein the method further comprises:
based on transmitting the first message:
monitoring the congestion status at the switch in the network; and
based on determining that the congestion status indicates that there is no longer congestion at the switch related to the data packets received at the switch from the source, transmitting a third message to the source requesting the source to send the data packets at a fifth rate that is faster than the second rate.

10. The computer program product of claim 8 , wherein the method further comprises:
based on transmitting the second message:
monitoring the congestion status at the switch in the network; and
based on determining that the congestion status indicates that there is no longer congestion at the switch related to the data requests received at the switch from the destination, transmitting a fourth message to the destination requesting the destination to send the data requests at a sixth rate that is faster than the fourth rate.

11. The computer program product of claim 8 , wherein the switch, the source, and the destination are OpenFlow switches and the network is an OpenFlow network.

12. The computer program product of claim 8 , wherein the switch is an OpenFlow switch.

13. The computer program product of claim 12 , wherein the forward congestion flag is set to indicate congestion at the switch related to the data packets received at the switch based on a queue at the switch that stores the data packets exceeding a specified capacity.

14. The computer program product of claim 12 , wherein the backward congestion flag is set to indicate congestion at the switch related to the data requests received at the switch based on a queue at the switch that stores the data requests exceeding a specified capacity.

Assignees

• IBM

Inventors

• Decusatis Casimer
• Fabrizio Petrini
• Krishnamurthy Rajaram B

Classifications

• H04L47/11 Primary

  Identifying congestion · CPC title

• H04L47/21

  using leaky-bucket · CPC title

• H04L47/215

  using token-bucket · CPC title

• H04L47/10

  Flow control; Congestion control · CPC title

• H04L47/12

  Avoiding congestion; Recovering from congestion · CPC title