US 6,581,166 B1 · Network fault detection and recovery

Abstract

A method of fault detection and recovery utilizes dual independent networks to provide fault-tolerance. These networks, a primary and alternate network, are utilized such that communications for a particular computer take place via that computer's preferred network by default. Faults are detected through the use of a heartbeat pinging mechanism to detect faults on the network itself and by periodic port integrity checks to detect port faults. The integrity of the non-default network and port are also periodically verified to assure effective fault recovery. Upon detection of a fault, a packet routing table in the detecting computer is altered to set the detecting computer's default network to the previously non-default network. Additionally, a new gateway packet is transmitted which allows other network computers to modify their routing tables to communicate with the detecting computer over its current default network.

Claims17

Dependent claims

A method of fault detection and recovery for use on a computing device that is communicatively connected to first and second networks via first and second standard network interfaces having first and second network ports respectively, and that is using an IP routing table for outgoing packet routing, the method comprising the steps of:

establishing the first network as a preferred network for receiving, via the first network port, incoming communications to the computing device;
transmitting a connectivity verification packet at a first connectivity verification rate to a connectivity verification server residing on the first network;
waiting for a first amount of time after transmitting the connectivity verification packet;
detecting that no reply for the connectivity verification packet is received before the expiration of the first amount of time, and in response registering a fault in the first network;
altering an entry in the IP routing table for outgoing packet routing such that the entry, which indicates the first network port as a default port, is changed to indicate the second network port as a default port and establishing a second connectivity verification rate for transmitting a connectivity verification packet to the connectivity verification server residing on the first network;
and
causing to be transmitted via the first and second network ports a new gateway packet announcing that incoming communications to the computing device should designate the second network port as a gateway to the computing device.

4 more elements

A method of fault detection and recovery for use on a computing device that is communicatively connected to first and second networks via first and second standard network interfaces having first and second network ports respectively, and that is using a port look-up table for outgoing packet routing, the method comprising the steps of:

establishing the first network as a preferred network for receiving, via the first network port, incoming communications to the computer;
transmitting a connectivity verification packet at a first connectivity verification rate to a connectivity verification server residing on the first network;
waiting for a first amount of time after transmitting the connectivity verification packet;
detecting that no reply for the connectivity verification packet is received before the expiration of the first amount of time, and in response registering a fault in the first network;
altering an entry in the port look-up table for outgoing packet routing such that the entry, which indicates the first network port as a default port, is changed to indicate the second network port as a default port;
causing to be transmitted via the second network port a new gateway packet announcing that incoming communications to the computing device should designate the second network port as a gateway to the computing device;
and
establishing a second connectivity verification rate for transmitting a connectivity verification packet to the connectivity verification server residing on the first network.

5 more elements

A computer-readable medium having computer-executable instructions for performing a method of fault detection and recovery for use on a computing device that is communicatively connected to first and second networks via first and second standard network interfaces having first and second network ports respectively, and that is using a IP routing table for outgoing packet routing, the method comprising the steps of:

establishing the first network as a preferred network for receiving, via the first network port, incoming communications to the computing device;
transmitting a connectivity verification packet at a first connectivity verification rate to a connectivity verification server residing on the first network;
waiting for a first amount of time after transmitting the connectivity verification packet;
detecting that no reply for the connectivity verification packet is received before the expiration of the first amount of time, and in response registering a fault in the first network;
altering an entry in the IP routing table for outgoing packet routing such that the entry, which indicates the first network port as a default port, is changed to indicate the second network port as a default port;
causing to be transmitted via the first and second network ports a new gateway packet announcing that incoming communications to the computing device should designate the second network port as a gateway to the computing device;
and
establishing a second connectivity verification rate for transmitting a connectivity verification packet to the connectivity verification server residing on the first network.

5 more elements

Dependencies are inferred from the claim text.

Network fault detection and recovery

Life of this patent

Network fault detection and recovery

Life of this patent

Abstract

Claims17