Methods and network device for oversubscription handling

Patent metadata

Field	Value
Publication number	US-9674102-B2
Application number	US-201414575855-A
Country	US
Kind code	B2
Filing date	Dec 18, 2014
Priority date	Dec 18, 2013
Publication date	Jun 6, 2017
Grant date	Jun 6, 2017

Abstract

Header information is extracted from a received packet by a data path portion of a network device. The data path portion is configured to buffer a data portion of received packets until the received packets are ready for transmission from the network device. The data path portion determines a first classification identifier for the received packet based on the header information. The data path portion determines a congestion state of the data path portion. The congestion state indicates a received packet rate of the first data path portion that exceeds a packet handling rate of a control path portion of the network device. The data path portion discards the first packet if the congestion state meets a discard threshold associated with the first classification identifier.

First claim

What is claimed is:

1. A method for processing packets in a network device, the method comprising:
extracting at least first header information from a first packet received by a first data path portion of the network device, the first data path portion configured to buffer a data portion of received packets until the received packets are ready for transmission from the network device;
determining, by the first data path portion, a first classification identifier for the first packet based on the first header information;
determining, by the first data path portion, a congestion state of the first data path portion, the congestion state indicating a received packet rate of the first data path portion that exceeds a packet handling rate of a control path portion of the network device, the control path portion being configured to process at least header information for the received packets to determine respective ports of the network device from which the received packets should be transmitted; and
discarding the first packet by the first data path portion, if the congestion state meets a discard threshold associated with the first classification identifier.

2. The method of claim 1 , wherein determining the first classification identifier comprises:
parsing at least one first header field from the first packet based on a predetermined packet protocol definition; and
selecting the first classification identifier for the first packet from a set of classification values based on a pattern match of the at least one first header field.

3. The method of claim 1 , further comprising scheduling the first header information for processing by the control path portion of the network device if the first packet is not discarded.

4. The method of claim 3 , wherein scheduling the first header information comprises:
parsing the first header information by a header parser to determine the first classification identifier;
sending the first header information, by the header parser, directly to a header scheduler of the first data path portion if the first classification identifier has a first value of a set of classification values; and
sending the first header information, by the header parser, to the header scheduler via a header buffer of the first data path portion if the first classification identifier has a second value of the set of classification values that corresponds to the header buffer.

5. The method of claim 4 , wherein sending the first header information to the header scheduler via the header buffer comprises selecting the header buffer from a plurality of header buffers, each header buffer of the plurality of header buffers corresponding to a classification value of the set of classification values.

6. The method of claim 5 , further comprising:
sending header information to the control path portion, by the header scheduler, from i) the plurality of header buffers and ii) the header parser based on a priority schedule for the set of classification values.

7. The method of 4 , wherein sending the first header information directly to the header scheduler comprises sending the first header information directly to the header scheduler if the first classification identifier has the first value of the set of classification values, the first value corresponding to a low latency packet classification.

8. The method of claim 4 , further comprising:
determining a current fill level of the header buffer as the congestion state; and
triggering a flow control message to be transmitted to a sending device if the current fill level meets a flow control threshold.

9. The method of claim 4 , further comprising:
determining a current fill level of the header buffer as the congestion state;
determining, by the control path portion, a second classification identifier for the first packet based on the first header information; and
causing the first packet to be discarded if the current fill level meets a discard threshold associated with the second classification identifier.

10. The method of claim 9 , wherein determining the second classification identifier comprises:
parsing at least one second header field from the first packet based on a reconfigurable packet protocol definition; and
selecting the second classification identifier for the first packet based on the at least one second header field parsed.

11. The method of claim 9 , wherein determining the second classification identifier comprises determining the second classification identifier by the control path portion based on flow characteristics associated with the first packet.

12. The method of claim 9 , determining the second classification identifier comprises determining the second classification identifier by the control path portion based on a content addressable memory lookup.

13. The method of claim 1 , further comprising adjusting the discard threshold based on the congestion state.

14. The method of claim 1 , wherein the congestion state is a congestion state of the first data path portion, the method further comprising:
extracting, by a second data path portion of the network device and for use by the control path portion, at least second header information from a second packet received by the second data path portion;
determining, by the second data path portion, a second classification identifier for the second packet based on the second header information;
determining, by the second data path portion, a congestion state of the second data path portion; and
selectively discarding the second packet if the congestion state of the second data path portion meets a discard threshold associated with the second classification identifier or sending the second header information from the second data path portion to the control path portion.

15. A network device, comprising:
a control path portion configured to process at least header information for received packets to determine respective ports of the network device from which the received packets should be transmitted; and
a first data path portion configured to:
buffer at least a data portion of the received packets until the received packets are ready for transmission from the network device,
extract first header information from a first packet received by the first data path portion,
determine a first classification identifier for the first packet based on the first header information,
determine a congestion state of the first data path portion, the congestion state indicating a received packet rate of the first data path portion that exceeds a packet handling rate of the control path portion, and
discard the first packet, if the congestion state meets a discard threshold associated with the first classification identifier.

16. The network device of claim 15 , the first data path portion comprising a parser configured to schedule the first header information for processing by the control path portion of the network device if the first packet is not discarded.

17. The network device of claim 16 , wherein the parser comprises:
a header scheduler, and
a header parser configured to i) parse the first header information, ii) send the first header information directly to the header scheduler if the first classification identifier has a first value of a set of classification values, and iii) send the first header information to the header scheduler via a header buffer of the parser if the first classification identifier has a second value of the set of classification values that corresponds to the header buffer.

18. The network

Assignees

• Marvell Israel (M I S L) Ltd
• Marvell Israel (M I S L ) Ltd

Inventors

• Levy Gil
• Kadosh Aviran

Classifications

• H04L49/9042

  Separate storage for different parts of the packet, e.g. header and payload · CPC title

• H04L47/32 Primary

  by discarding or delaying data units, e.g. packets or frames · CPC title

• H04L69/12

  Protocol engines · CPC title

• H04L47/24

  Traffic characterised by specific attributes, e.g. priority or QoS · CPC title

• H04L47/11

  Identifying congestion · CPC title