Browse Prior Art Database

Method For Failure Processing During Dynamic Route Creation

IP.com Disclosure Number: IPCOM000048486D
Original Publication Date: 1982-Feb-01
Included in the Prior Art Database: 2005-Feb-08
Document File: 3 page(s) / 69K

Publishing Venue

IBM

Related People

George, FD: AUTHOR [+2]

Abstract

A procedure is described for maintaining consistent source-independent explicit routing tables while network failures are concurrent with route creation procedures. This method, called a purge filter, is needed to make the protocol of the preceding article fail-safe. See the preceding article for a brief description of source-independent explicit routing.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 3

Method For Failure Processing During Dynamic Route Creation

A procedure is described for maintaining consistent source-independent explicit routing tables while network failures are concurrent with route creation procedures. This method, called a purge filter, is needed to make the protocol of the preceding article fail-safe. See the preceding article for a brief description of source-independent explicit routing.

When a TG, TG1 fails in an explicit routing environment, and route A.i goes over TG1, then all nodes that use A.i through TG1 are informed by a message ER INOP. In a dynamic routing environment, if a node adjacent to A.i had a route A.i in its RT, and A.i.TG1equals1, then it sets A.i.TG1equals0. It also generates an ER DEACT REPLY (A.i), and deletes the RT entry if no other sessions use a portion of A.i through that node from other sources besides TG1.

The fact that ER SETUP activates an ER towards the ER source (see the preceding article) has an important impact on the processing of ER INOP which occurs while processing ER SETUP. The basic operation at failure remains the same: ER INOP sets all flags off, and a forward flowing ER DEACT REPLY is generated. (see original)

Fig. 2 illustrates an ER SETUP REPLY crossing an ER INOP. All flags for the route to A are set off, either by ER INOP or by ER DEACT REPLY, but some flags for the route to D are left on. A technique is needed to insure that the flags for the route to node D are also turned off.

Fig. 2 illustrates a technique called a purge filter operating on ER SETUP REPLY. The purge filter prevents the ER SETUP REPLY from reaching nodes A, B, and C. To restore the flags already set by ER SETUP REPLY, the purge filter changes the ER SETUP REPLY into an ER DEACT REPLY. (see original)

The purge filter works as follows: After node C propagates the ER INOP, it rejects all relevant ER SETUP REPLYs that it receives until node E acknowledges that the ER INOP has been received. This is called a purge filter, as node C filters the incoming traffic and "purges" certain messages. After node E's acknowledgement, node C again will process ER SETUP REPLYs. The situation in Fig. 1 is prevented since node E erects its own p...