Detecting intermittent network link failures

• US 8,917,610 B2
• Filed: 07/12/2010
• Issued: 12/23/2014
• Est. Priority Date: 07/12/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Astigarraga, Tara, Dickens, Louie A., Winarski, Daniel J.
• Primary Examiner(s): Sam, Phirin
• Assistant Examiner(s): Hsiung, Hai-Chang
• Application Number:
  US12/834,599
• Publication Number:
  US 20120008506A1
• Time in Patent Office: 1,625 Days
• Field of Search: 370/242
• US Class Current: 370/242
• CPC Class Codes: H04L 41/0677 Localisation of faults
• CPC Class Codes: H04L 43/0811 by checking connectivity
• CPC Class Codes: H04L 43/10 Active monitoring, e.g. heartbeat, ping or trace-route

Abstract

An apparatus, system, and method are disclosed for detecting intermittent network link failures. A tracking module tracks link failures of a network link of a network over a specified time interval. A failure module determines the network link is failing in response to a number of link failures exceeding a specified failure threshold. A mitigation module mitigates communications over the network link.

Claims

1. An apparatus for detecting intermittent network link failures, the apparatus comprising:

a non-tangible computer readable storage medium storing a computer readable program;

a processor executing the computer readable program, the computer readable program comprising;

a detection module detecting link failures of a network link of a network by notifying an Ethernet data link layer of each link failure from an Ethernet physical layer, the Ethernet data link layer notifying a network agent residing at the data link layer of the link failure, and the network agent logging the link failure;

a tracking module tracking the link failures over a specified time interval, wherein a link failure comprises missing a heartbeat message during a heartbeat time interval;

a failure module determining the network link is failing in response to a number of link failures during the specified time interval exceeding a specified failure threshold, wherein the specified failure threshold t is calculated from a service level as t=ksn wherein k is a non-zero constant, s is the service level, and n is a number of heartbeat messages during the specified time interval; and

a mitigation module mitigating communications over the network link by removing the failing network link from the network.

2. The apparatus of claim 1, wherein the specified failure threshold is zero link failures per the specified time interval.

3. The apparatus of claim 1, wherein the network employs a spanning tree protocol to organize a loop-free topology.

4. A computer program product comprising a non-tangible computer readable storage medium storing a computer readable program executing on a processor to perform operations for detecting intermittent network link failures, the operations of the computer program product comprising:

tracking link failures of a network link of a network over a specified time interval by notifying an Ethernet data link layer of each link failure from an Ethernet physical layer, the Ethernet data link layer notifying a network agent residing at the data link layer of the link failure, and the network agent logging the link failure, wherein a link failure comprises missing a heartbeat message during a heartbeat time interval;

determining the network link is failing in response to a number of link failures during the specified time interval exceeding a specified failure threshold, wherein the specified failure threshold t is calculated from a service level as t =ksn wherein k is a non-zero constant, s is the service level, and n is a number of heartbeat messages during the specified time interval; and

mitigating communications over the network link by removing the failing network link from the network.

5. The computer program product of claim 4, wherein the specified failure threshold is zero link failures per the specified time interval.

6. The computer program product of claim 4, further comprising detecting each link failure.

7. The computer program product of claim 4, wherein the network agent is notified using an internet control message protocol message.

8. The computer program product of claim 4, further comprising communicating a removal notice to remove the network link from the network.

9. A method for detecting intermittent network link failures, the method comprising:

tracking link failures of a network link of a network over a specified time interval by notifying an Ethernet data link layer of each link failure from an Ethernet physical layer, the Ethernet data link layer notifying a network agent residing at the data link layer of the link failure, and the network agent ligging the link failure, wherein a link failure comprises missing a heartbeat message during a heartbeat time interval;

determining the network link is failing in response to a number of link failures during the specified time interval exceeding a specified failure threshold, wherein the specified failure threshold t is calculated from a service level as t =ksn wherein k is a non-zero constant, s is the service level, and n is a number of heartbeat messages during the specified time interval; and

mitigating communications over the network link by removing the failing network link from the network.

10. The method of claim 9, wherein the specified failure threshold is zero link failures per the specified time interval.

11. A system to detect intermittent network link failures, the system comprising:

a network;

a data processing device in communication with network, the data processing device comprising;

a tracking module tracking link failures of a network link of the network over a specified time interval by notifying an Ethernet data link layer of each link failure from an Ethernet physical layer, the Ethernet data link layer notifying a network agent residing at the data link layer of the link failure, and the network agent logging the link failure,

wherein a link failure comprises missing a heartbeat message during a heartbeat time interval;

a failure module determining the network link is failing in response to a number of link failures during the specified time interval exceeding a specified failure threshold, wherein the specified failure threshold t is calculated from a service level as t =ksn wherein k is a non-zero constant, s is the service level, and n is a number of heartbeat messages during the specified time interval; and

a mitigation module mitigating communications over the network link by removing the failing network link from the network.

12. The system of claim 11, wherein the network is an Ethernet network.

13. The system of claim 12, wherein the Ethernet network is selected from a group consisting of a one Gigabit Fiber Channel over Ethernet (FCoE) network, a ten Gigabit FCoE network, an Internet Small Computer System Interface (iSCSI) network, a Hyper Small Computer System Interface (HyperSCSI) network, and an Advanced Technology Attachment (ATA) over Ethernet (AoE) network.

14. The system of claim 12, wherein the Ethernet network employs a transport medium selected from the group consisting of a fiber optic transport medium, a copper transport medium, and a wireless transport medium.