Preventing transmission of errors in a computing network

• US 10,397,143 B1
• Filed: 06/19/2017
• Issued: 08/27/2019
• Est. Priority Date: 06/19/2017
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Plenderleith, Jamie, Judge, Alan Michael
• Primary Examiner(s): Henson, Jamaal
• Application Number:
  US15/627,080
• Time in Patent Office: 799 Days
• Field of Search: None
• CPC Class Codes: H04L 1/0061 Error detection codes
• CPC Class Codes: H04L 1/20 using signal quality detector
• CPC Class Codes: H04L 1/24 Testing correct operation
• CPC Class Codes: H04L 49/251 Cut-through or wormhole routing
• CPC Class Codes: H04L 49/257 Cut-through or wormhole routing
• CPC Class Codes: H04L 49/55 Prevention, detection or correction of errors
• CPC Class Codes: H04L 49/552 by ensuring the integrity of packets received through redundant connections
• CPC Class Codes: H04L 63/12 Applying verification of the received information cryptographic mechanisms or cryptographic arrangements for data integrity or data verification H04L9/32

Abstract

Technologies are provided for preventing transmission of cyclic redundancy check (CRC) errors, and/or other errors, within a computer network that uses cut-through forwarding of network packets. For example, counts of CRC errors can be maintained for network packets processed by network devices. When a CRC error condition is detected based on the counts of CRC errors, cut-through forwarding is disabled on at least a portion of the network (e.g., on one or more specific network ports and/or on one or more specific network devices). While cut-through forwarding is disabled, the affected portion of the network uses store-and-forward for processing network packets. When the CRC error condition no longer exists, then cut-through forwarding is re-enabled on the portion of the network.

Claims

1. A method, implemented by a network device, for preventing transmission of cyclic redundancy check (CRC) errors, the method comprising:

maintaining counts of CRC errors for network packets processed by network ports of the network device, wherein the network device supports cut-through forwarding, and wherein cut-through forwarding is enabled on the network ports;

upon detecting a CRC error condition indicated by CRC errors of the network packets increasing above an error threshold;

if the network device supports tracking outbound CRC errors for transmitted network packets;

detecting the CRC error condition for a particular network port of the network device; and

disabling cut-through forwarding on the particular network port so that the particular network port uses store-and-forward processing when processing network packets, wherein disabling cut-through forwarding on the particular network port does not affect whether the other network ports of the network device use cut-through forwarding; and

while cut-through forwarding is disabled on the particular network port, processing network packets via the particular network port using store-and-forward processing; and

otherwise, if the network device does not support tracking outbound CRC errors for transmitted network packets;

detecting the CRC error condition;

disabling cut-through forwarding for all network ports of the network device; and

while cut-through forwarding is disabled on all of the network ports, processing network packets via all of the network ports using store-and-forward network packet processing.

2. The method of claim 1, further comprising:

upon detecting that the CRC error condition has ended for the particular network port;

enabling cut-through forwarding on the particular network port so that the particular network port uses cut-through forwarding when processing network packets;

wherein the CRC error condition has ended when no inbound network packets are received with CRC errors for a period of time.

3. The method of claim 1, further comprising:

upon detecting that the CRC error condition has ended for the particular network port;

enabling cut-through forwarding on the particular network port so that the particular network port uses cut-through forwarding when processing network packets;

wherein the CRC error condition has ended when a rate of CRC errors has dropped below a second error threshold.

4. The method of claim 1, wherein the error threshold is based upon a percentage of network packets received, transmitted, or received and transmitted.

5. The method of claim 1, wherein maintaining counts of CRC errors comprises maintaining a count of inbound CRC errors and a count of outbound CRC errors for each of the network ports of the network device, and wherein the particular network port is transmitting or receiving the network packets.

6. A network device comprising:

a processing unit; and

network ports;

wherein the network device performs operations, the operations comprising;

maintaining counts of packet errors for network packets processed by the network ports, wherein the packet errors comprise cyclic redundancy check (CRC) errors;

upon detecting a packet error condition based on the counts of packet errors;

if the network device supports tracking outbound CRC errors for transmitted network packets;

detecting the packet error condition for one of the network ports; and

disabling cut-through forwarding on only the one network port; and

while cut-through forwarding is disabled on the one network port, processing network packets via the one network port using store-and-forward network packet processing; and

otherwise, if the network device does not support tracking outbound CRC errors for transmitted network packets;

detecting the packet error condition; and

disabling cut-through forwarding for all network ports of the network device; and

while cut-through forwarding is disabled on all of the network ports, processing network packets via all of the network ports using store-and-forward network packet processing.

7. The network device of claim 6, wherein the network device does not support tracking outbound CRC errors for transmitted network packets, and wherein counts of packet errors are maintained for inbound network packets.

8. The network device of claim 7, wherein the packet error condition is detected when a rate of packet errors exceeds an error threshold.

9. The network device of claim 7, the operations further comprising:

upon detecting that the packet error condition has ended;

enabling cut-through forwarding on all network ports of the network device.

10. The network device of claim 6, wherein the network device supports tracking outbound CRC errors for transmitted network packets, and wherein maintaining counts of packet errors comprises maintaining a count of inbound packet errors and a count of outbound packet errors for each of the network ports of the network device.

11. The network device of claim 10, wherein the packet error condition is detected when a rate of outbound packet errors for the one network port exceeds an error threshold.

12. The network device of claim 10, the operations further comprising:

upon detecting that the packet error condition has ended for the one network port;

enabling cut-through forwarding on the one network port so that the one network port uses cut-through forwarding when transmitting network packets.

13. The network device of claim 6, wherein the packet errors comprise header errors, forward error correction errors, or framing errors.

14. The network device of claim 6, wherein the network device is a network switch.

15. The network device of claim 6, wherein the packet error condition is detected based on counts of inbound packet errors.

16. A computer-readable storage medium storing computer-executable instructions for causing a computing device to perform operations, the operations comprising:

obtaining cyclic redundancy check (CRC) error information for network packets processed by a network device, wherein the network device supports cut-through forwarding, and wherein cut-through forwarding is enabled on a plurality of network ports of the network device;

based on the obtained CRC error information, detecting a CRC error condition associated with the network device; and

if the network device supports tracking outbound CRC errors for transmitted network packets;

detecting the CRC error condition for one of the network ports; and

in response to detecting the CRC error condition, sending a command to the network device to disable cut-through forwarding on only the one network port so that the one network port uses store-and-forward when processing network packets; and

otherwise, if the network device does not support tracking outbound CRC errors for transmitted network packets;

detecting the CRC error condition; and

in response to detecting the CRC error condition, sending a command to the network device to disable cut-through forwarding for all network ports of the network device so that all of the network ports process network packets using store-and-forward network packet processing.

17. The computer-readable storage medium of claim 16, wherein the CRC error information is obtained by polling the network device to collect counts of inbound CRC errors for received network packets processed by the network device.

18. The computer-readable storage medium of claim 16, the operations further comprising:

upon detecting that the CRC error condition has ended;

sending a command to the network device to enable cut-through forwarding on the network port or ports that were disabled.

19. The computer-readable storage medium of claim 16, wherein the CRC error information is obtained via an alert received, by a monitoring service, from the network device.