Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Fair Assignment of Time-Limited Resources

IP.com Disclosure Number: IPCOM000111363D
Original Publication Date: 1994-Feb-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 2 page(s) / 36K

Publishing Venue

IBM

Related People

Mitchell, J: AUTHOR [+2]

Abstract

Described is a process for providing fair allocation of time-limited resources. When a client's allotted time with the resource is up and no other clients are requesting the resource, the resource is not taken back, but a log is kept of how long this resource has been held by this client. When a new client requests a resource and none is available, the log is consulted for the client which has held its resource for the longest time. If the resource held the longest is past its limit, that resource is taken away from the owning client and given to the new client. If no client has exceeded its allotted time of use, then the new client is put into a queue and will be notified when a resource becomes available.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 85% of the total text.

Fair Assignment of Time-Limited Resources

      Described is a process for providing fair allocation of
time-limited resources.  When a client's allotted time with the
resource is up and no other clients are requesting the resource, the
resource is not taken back, but a log is kept of how long this
resource has been held by this client.  When a new client requests a
resource and none is available, the log is consulted for the client
which has held its resource for the longest time.  If the resource
held the longest is past its limit, that resource is taken away from
the owning client and given to the new client.  If no client has
exceeded its allotted time of use, then the new client is put into a
queue and will be notified when a resource becomes available.  The
next time a client exceeds its limit, the resource is taken away from
that client and given to the client who has been waiting on the
resource the longest.  If a using client voluntarily gives up its
resource, the resource will also be given to any client that is
waiting on the resource.

      In this way, resources are only taken from a client when they
are needed by some other client.  This saves the cost of needlessly
de-allocating the resource, and it also ensures that the resource
cannot be monopolized by a client.  It also ensures that when a
resource is needed, it is taken from the client that has held the
resource the longest, which leads to a fairer allocation scheme.

      Previously...