Browse Prior Art Database

Variable Sized Refresh Intervals for Network Connections

IP.com Disclosure Number: IPCOM000112541D
Original Publication Date: 1994-May-01
Included in the Prior Art Database: 2005-Mar-27
Document File: 2 page(s) / 85K

Publishing Venue

IBM

Related People

Drake, JE: AUTHOR [+8]

Abstract

Disclosed is a process for maintaining the reserved resources in a computer network that reserves resources. The refresh messages indicate to the nodes participating in the network connection that the resources allocated to that network connection are still required. In the event of intermediate node failure, the refresh messages carry enough information for a intermediate node to re-reserve the resources for the network connection. In the event of end node failure, the absence of refresh messages indicates to the intermediate nodes that the resources may be reallocated for other network connections.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Variable Sized Refresh Intervals for Network Connections

      Disclosed is a process for maintaining the reserved resources
in a computer network that reserves resources.  The refresh messages
indicate to the nodes participating in the network connection that
the resources allocated to that network connection are still
required.  In the event of intermediate node failure, the refresh
messages carry enough information for a intermediate node to
re-reserve the resources for the network connection.  In the event of
end node failure, the absence of refresh messages indicates to the
intermediate nodes that the resources may be reallocated for other
network connections.

      The problem of selecting an interval for the refresh messages
is that a tradeoff must be made between the rate at which the refresh
messages can be processed by the participating nodes (which favors a
longer interval) and the responsiveness with which the network
recovers resources in the event of failures (which favors a shorter
interval.)  Furthermore, if network connections are aggregated at
intermediate nodes, this process needs to provide a method to account
for the resources when different refresh intervals for the aggregated
small network connections are used.

      There is a network wide minimum refresh interval denoted T sub
MinRefresh .  All refresh intervals are positive, non-zero, integer
multiples of T sub MinRefresh .  There is also a network wide maximum
refresh interval denoted T sub MaxRefresh .  This interval is also a
multiple of T sub MinRefresh .  T sub MinRefresh and T sub MaxRefresh
are expressed in some suitable time units.

1.  When the origin node sends the connection setup request, the
    default refresh interval is T sub MaxRefresh  for all individual
    paths associated with the network connection.  (Point-to-point
    network connections have a forward path and may have a return
    path.  Point-to-multi-point network connections have one or more
    forward paths, and multi-point network connections have one or
    more forward and return paths.)

2.  The recipients of the connection setup message for each path
    (each intermediate node and the destination node)  place the
    minimum refresh interval they are willing to accept for the path
    in the connection setup reply.  T sub MinRefresh_i  denotes the
    minimum acceptable refresh interval where i  indexes the
    intermediate nodes and destination node on a path.

3.  The origin node choos...