Congestion mitigation and avoidance

Patent metadata

Field	Value
Publication number	US-9350665-B2
Application number	US-201213601957-A
Country	US
Kind code	B2
Filing date	Aug 31, 2012
Priority date	Aug 31, 2012
Publication date	May 24, 2016
Grant date	May 24, 2016

Abstract

In one embodiment, a method is provided for congestion mitigation. In this method, a congestion level of data flows along a path is monitored relative to a threshold. Here, each data flow is associated with a flow identifier. Based on detection that the congestion level exceeds the threshold, the flow identifier of a number (e.g., one or more) of the data flows is changed to a different flow identifier. This change results in the number of the data flows redirected to a different path. In an alternate embodiment, another method is provided for congestion avoidance. In this method, a data flow transmitted along a single path is received, and the data flow is split into multiple micro data flows. A different flow identifier is then assigned to each micro data flow. This change results in the multiple micro data flows distributed along multiple different paths.

First claim

What is claimed is:

1. A method comprising:
identifying, at a source device, a plurality of equal cost paths from the source device;
establishing, at the source device, a data flow along a path of the plurality of equal cost paths and associating the data flow with a flow identifier, wherein the flow identifier is based on data associated with a first Open System Interconnection (OSI) model layer;
monitoring, at the source device, a congestion level of the data flow along the path of the plurality of equal cost paths;
in response to detecting that the congestion level exceeds a threshold, splitting the data flow into a first split flow and a second split flow;
sending the first split flow along a first split flow path and sending the second split flow along a second split flow path;
wherein splitting the data flow comprises splitting the data flow based on data associated with a second OSI model layer, wherein the second OSI model layer is higher in the OSI model than the first OSI model layer,
wherein sending the first split flow along the first split flow path and sending the second split flow along the second split flow path comprises changing a flow identifier for at least one of the first split flow or the second split flow, and
wherein the data flow is identified in OSI layer 3 based upon a combination of an Internet Protocol source address and an Internet Protocol destination address and the data flow is split at OSI layer 4 into the first split flow and the second split flow based on at least one of a User Datagram Protocol source port or a User Datagram Protocol destination port.

2. The method of claim 1 , further comprising inserting a start marker at a beginning of transmission of the first split flow, the start marker indicating a start of the transmission of the first split flow.

3. The method of claim 2 , wherein the start marker is a data packet that includes source information, destination information, and the different flow identifier.

4. The method of claim 1 , further comprising inserting an end marker at an end of transmission of the first split flow, the end marker indicating the end of the transmission of the first split flow.

5. The method of claim 4 , wherein the end marker is a data packet that includes source information, destination information, and the flow identifier.

6. The method of claim 1 , wherein the monitoring of the congestion level comprises receiving a congestion notification message along the path of the plurality of equal cost paths, the congestion notification message including information about an extent of congestion along the path.

7. The method of claim 1 , wherein the flow identifier consists of values associated with layer 2 of the Open System Interconnection (OSI) model and the split flow identifier comprises values associated with layer 3 of the OSI model.

8. Logic encoded on one or more non-transitory, tangible media and when executed cause operations to be performed, the operations comprising:
identifying a plurality of equal cost paths from the source device;
establishing a data flow along a path of the plurality of equal cost paths and associating the data flow with a flow identifier, wherein the flow identifier is based on data associated with a first Open System Interconnection (OSI) model layer;
monitoring a congestion level of the data flow along the path of the plurality of equal cost paths;
in response to detecting that the congestion level exceeds a threshold, splitting the data flow into a first split flow and a second split flow;
sending the first split flow along a first split flow path and sending the second split flow along a second split flow path;
wherein splitting the data flow comprises splitting the data flow based on data associated with a second OSI model layer, wherein the second OSI model layer is higher in the OSI model than the first OSI model layer,
wherein sending the first split flow along the first split flow path and sending the second split flow along the second split flow path comprises changing a flow identifier for at least one of the first split flow or the second split flow, and
wherein the data flow is identified in OSI layer 3 based upon a combination of an Internet Protocol source address and an Internet Protocol destination address and the data flow is split at OSI layer 4 into the first split flow and the second split flow based on at least one of a User Datagram Protocol source port or a User Datagram Protocol destination port.

9. The logic of claim 8 , the operations further comprising inserting a start marker at a beginning of transmission of the first split flow, the start marker indicating a start of the transmission of the first split flow.

10. The logic of claim 9 , wherein the start marker is a data packet that includes source information, destination information, and the different flow identifier.

11. The logic of claim 8 , the operations further comprising inserting an end marker at an end of transmission of the first split flow, the end marker indicating the end of the transmission of the first split flow.

12. The logic of claim 11 , wherein the end marker is a data packet that includes source information, destination information, and the flow identifier.

13. The logic of claim 8 , wherein the operation of monitoring the congestion level comprises receiving a congestion notification message along the path of the plurality of equal cost paths, the congestion notification message including information about an extent of congestion along the path of the plurality of equal cost paths.

14. An apparatus comprising:
a congestion mitigation module located at a source device and comprising one or more processors and having instructions that when executed by the one or more processors cause the one or more processors to:
identify a plurality of equal cost paths from the source device;
establish a data flow along a path of the plurality of equal cost paths and associate the data flow with a flow identifier, wherein the flow identifier is based on data associated with a first Open System Interconnection (OSI) model layer;
monitor a congestion level of the data flow along the path of the plurality of equal cost paths, the data flow being associated with a flow identifier; and
in response to detecting that the congestion level exceeds a threshold, split the data flow into a first split flow and a second split flow; and
send the first split flow along a first split flow path and send the second split flow along a second split flow path;
wherein the instructions that cause the processor to split the data flow comprise instructions that cause the processor to split the data flow based on data associated with a second OSI model layer, wherein the second OSI model layer is higher in the OSI model than the first OSI model layer,
wherein the instructions that cause the processor to send the data flow comprise instructions that cause the processor to send the first split flow and send the second split flow comprise instructions that change a flow identifier for at least one of the first split flow or the second split flow, and
wherein the data flow is identified in OSI layer 3 based upon a combination of an Internet Protocol source address and an Internet Protocol destination address and the data flow is split at OSI layer 4 into the first split flow and the second split flow based on at least one of a User Datagram Protocol source port or a User Datagram Protocol destination port.

15. The apparatus of claim 14 , wherein the instructions further cause the one or more processors to insert a start marker at a beginni

Assignees

• Yeung Derek Man-Kit
• Sajassi Ali
• Zhang Hong
• Cisco Tech Inc

Inventors

• Yeung Derek Man-Kit
• Sajassi Ali
• Zhang Hong

Classifications

• H04L47/125 Primary

  by balancing the load, e.g. traffic engineering · CPC title