Detecting sources of computer network failures

Patent metadata

Field	Value
Publication number	US-10601644-B2
Application number	US-201715809836-A
Country	US
Kind code	B2
Filing date	Nov 10, 2017
Priority date	Nov 10, 2017
Publication date	Mar 24, 2020
Grant date	Mar 24, 2020

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for detecting sources of computer network failures. One of the methods includes identifying a network flow in a computer network between a source and a destination; performing a first probe to determine whether there is end-to-end connectivity between the source and the destination; in response to determining that there is no end-to-end connectivity between the host and the destination, performing one or more additional probes including a second probe to determine whether each hop in the path of the network flow between the source and the destination is operational including requesting that the source transmit a respective first trace diagnostic packet to each hop in the path of the network flow; and determining whether at least one link of the computer network that is part of the path of the network flow has failed based on the results.

First claim

What is claimed is:

1. A method comprising:
identifying a network flow of packets in a computer network between a source and a destination, wherein the network flow corresponds to a particular path in the network between the source and the destination;
performing a first probe to determine whether there is end-to-end connectivity between the source and the destination including requesting that the source transmit an end-to-end diagnostic packet to the destination, wherein the end-to-end diagnostic packet includes information comprising a source identifier field and a destination identifier field that include respective identifiers for the source and destination, respectively, that match the packets in the network flow between the source and destination such that one or more network devices in the computer network forward the end-to-end diagnostic packet on the path of the network flow;
in response to determining that there is no end-to-end connectivity between the host and the destination, performing one or more additional probes including a second probe to determine whether each hop, of a plurality of hops, in the path of the network flow between the source and the destination is operational including requesting that the source transmit a respective first trace diagnostic packet to each hop in the path of the network flow, wherein each first trace diagnostic packet includes information that causes a particular hop to send a first trace response packet responsive to the first trace diagnostic packet to the source; and
determining whether at least one link of the computer network that is part of the path of the network flow has failed based on the results of the first probe and the one or more additional probes.

2. The method of claim 1 , wherein performing the one or more additional probes comprises performing a third probe to determine whether each hop in a path between the destination and the source is operational including requesting that the destination transmit a respective second trace diagnostic packet to each hop in the path between the destination and the source, each trace diagnostic packet having information that cause the respective hop to send a second trace response packet responsive to the second trace diagnostic packet to the source host.

3. The method of claim 2 , wherein the second trace diagnostic packet associated with a respective hop has a particular value in a time-to-live field of the respective second trace diagnostic packet such that the second trace diagnostic packet will expire at a corresponding hop causing the hop to send a time exceeded message to the host in response to that second trace diagnostic packet, wherein the time exceeded message includes a portion of the second trace diagnostic packet that includes an identifier of the packet.

4. The method of claim 2 , further comprising:
generating probe result information that include results of the first probe, the second probe, and the third probe;
analyzing the probe results to determine a visit count and a failure count for each link in the network, the visit count for a corresponding link indicating a number of times that packets have traveled the link and the failure count for a corresponding link indicating a number of times that the link has shown signs of failure; and
generating a graph of the computer network, the graph including edges that each correspond to a respective link in the computer network and weights for each edge that are determined based on at least one of the visit count and the failure count for the link corresponding to the edge.

5. The method of claim 4 , further comprising analyzing the graph to detect at least one link in the computer link that has likely failed.

6. The method of claim 1 , wherein the end-to-end diagnostic packet has specified values in one or more particular fields to indicate that the end-to-end diagnostic packet is a diagnostic packet.

7. The method of claim 1 , wherein the diagnostic packet is a Transport Control Protocol packet, and wherein the specified values comprise zero values for a flags field.

8. The method of claim 1 , wherein the first trace diagnostic packet associated with a respective hop has a particular value in a time-to-live field of the respective first trace diagnostic packet such that the first trace diagnostic packet will expire at a corresponding hop causing the hop to send a time exceeded message to the host in response to that first trace diagnostic packet, wherein the time exceeded message includes a portion of the second trace diagnostic packet that includes an identifier of the packet.

9. The method of claim 8 , wherein the identifier is placed in a sequence number of a TCP packet or in a length field of a UDP packet.

10. The method of claim 1 , wherein performing the first probe further comprises:
determining whether the source has received an end-to-end response packet responsive to the end-to-end diagnostic packet;
in response to determining that the source has received the end-to-end response packet, determining that there is end-to-end connectivity between the source and the destination; and
in response to determining that the source has not received the end-to-end response packet, determining that there is no end-to-end connectivity between the source and the destination.

11. The method of claim 1 , further comprising:
determining whether the source has received the first trace response packet from a particular hop;
in response to determining that the source has received the first trace response packet from a particular hop, determining that the particular hop is operational; and
in response to determining that the source has not received the first trace response packet from a particular hop, determining that the particular hop is not operational.

12. The method of claim 1 , wherein identifying the network flow comprises:
obtaining retransmission information from one or more hosts in the computer network;
analyzing the re-transmission information to detect one or more network flows;
for each network flow of the one or more network flows, determine a re-transmission count from the re-transmission information;
identify a group of the one or more network flows whose re-transmission count exceeds a threshold; and
selecting the network flow from the group.

13. The method of claim 12 , wherein identifying the network flow further comprises:
for each network flow in the group, detecting if a destination of the network flow has failed; and
updating the group to exclude any network flow whose corresponding destination has failed.

14. A system comprising:
one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
identifying a network flow of packets in a computer network between a source and a destination, wherein the network flow corresponds to a particular path in the network between the source and the destination;
performing a first probe to determine whether there is end-to-end connectivity between the source and the destination including requesting that the source transmit an end-to-end diagnostic packet to the destination, wherein the end-to-end diagnostic packet includes information comprising a source identifier field and a destination identifier field that include respective identifiers for the source and destination, respectively, that match the packets in the network flow between the source and destination such that one or more network devices in the computer network forward the end-to-end diagnosti

Assignees

• Twitter Inc

Inventors

• Shaw Anatole
• Roberts Anthony
• Demir Berk D
• Chen Christopher

Classifications

• H04L49/25

  Routing or path finding in a switch fabric · CPC title

• H04L43/0811

  by checking connectivity · CPC title

• H04L43/10

  Active monitoring, e.g. heartbeat, ping or trace-route · CPC title

• H04L41/0695 Primary

  the faulty arrangement being the maintenance, administration or management system · CPC title

• H04L41/0677

  Localisation of faults · CPC title