Peer-to-peer fault detection

• US 20030018930A1
• Filed: 07/18/2001
• Published: 01/23/2003
• Est. Priority Date: 07/18/2001
• Status: Active Grant

• Current Assignee: Oscar Mora, Roberto Ponticelli, Roger Pinate
• Original Assignee: Oscar Mora, Roberto Ponticelli, Roger Pinate
• Inventors: Pinate, Roger, Mora, Oscar, Ponticelli, Roberto
• Granted Patent: US 6,880,100 B2
• Time in Patent Office: Days
• US Class Current: 714/42
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: H04L 67/104 Peer-to-peer [P2P] networks
• CPC Class Codes: H04L 67/1048 Departure or maintenance mechanisms
• CPC Class Codes: H04L 69/329 in the application layer [OSI layer 7]
• CPC Class Codes: H04L 69/40 for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection management of faults, events, alarms or notifications in data switching networks H04L41/06

Abstract

This inventions discloses a fault or abnormal status detection mechanism applied to distributed networks considering a shared media. A set of programmed routines executing in every node connected to a distributed network establish pairs of inter-monitoring nodes. At system power up, every node searches for another free, unpaired node using random timeout messaging generation to prevent collisions of network messages. After all pairs have been formed, a periodic check message is sent from one node to its associated partner (and vice versa) requesting its operating status. A response, containing a node'"'"'s status, is generated upon receipt of a check message. If an abnormal status is received, or there is no response, a report is sent to a monitoring station or stored as a failure network event. The timeout period, in the order of seconds, between each check message is adjusted according to the number of connected nodes and available network bandwidth. This prevents excessive traffic generation that could affect performance. A longer timeout period, in the order of minutes, indicates when to restart the pair assignation process again, periodically changing all established pairs.

Claims

1. A method for for peer-to-peer fault detection for distributed, shared media networks, the method comprising the steps of:

having a distributed network comprising a group of nodes sharing a common communication network supporting point-to-point and broadcast messaging;

Using peer-to-peer logic fault detection in which each node finds a pair to establish a peer-to-peer checking mechanism at network power up; and

Reporting any irregular or non-received responses.

2. The method in claim 1 which includes the following steps comprising:

Using random or pseudo-random timeout generation along with broadcast messaging, the nodes are pair up until every node has an associated pair;

Assigning a node a timeout period to periodically send a status report message to its associated partner node;

Having the partner node generate and send a status report back; and

Reporting any irregular status or non-received response.

3. The method in claim 1 which includes any node having the ability to request tor reestablish of network peer-to-peer pairs

4. The method in claim 1 which includes the following steps comprising:

Having a Node send a check message to its paired Node requesting its updated status;

Having its paired Node reply with a reply check message with any abnormal status;

Resending check message if no reply check message is received;

Repeating previous step a set number of times; and

Reporting any non-received response if no reply check message is received.

5. The method in claim 1 which includes the following steps comprising:

Having a node that does not have a node pair send a status report to a paired node; and

Having the paired node report any irregular status but not send a reply check message to the node.

10. The device in claim 1 further comprising:

Having a node that does not have a node pair send a status report to a paired node; and

Having the paired node report any irregular status but not send a reply check message.

6. A network device for for peer-to-peer fault detection for distributed, shared media networks comprising:

A distributed network comprising a group of nodes sharing a common communication network supporting point-to-point and broadcast messaging;

A fault detection means using peer-to-peer logic fault detection in which each node finds a pair to establish a peer-to-peer checking mechanism at network power up; and

A reporting means reporting any irregular or non-received responses.

7. The device in claim 6 further comprising:

Having a random or pseudo-random timeout generation means along with broadcast messaging means to pair up nodes until every node has an associated pair;

Having the node is assigned a timeout period to periodically send a status report message to its associated partner node;

Having the partner node generate and send a status report back to the first node; and

Reporting an irregular status or non-received response.

8. The device in claim 6 further comprising any node being able to request to reestablish of network peer-to-peer pairs

9. The device in claim 6 further comprising:

Having a Node send a check message to its paired Node requesting its updated status,

Having its paired Node reply with a reply check message with any abnormal status,

Resending check message if no reply check message is received;

Repeating previous step a set number of times; and

Reporting any irregular or non-received response.