Browse Prior Art Database

Intelligent Distributed System

IP.com Disclosure Number: IPCOM000109790D
Original Publication Date: 1992-Sep-01
Included in the Prior Art Database: 2005-Mar-24
Document File: 3 page(s) / 83K

Publishing Venue

IBM

Related People

Hatta, T: AUTHOR [+3]

Abstract

Described herein is a control method for improving the reliability of distributed computer systems by shutting down the power of unnecessary machines and for increasing the availability of systems by providing the fault tolerance capability. The technique is widely applicable to all distributed computer systems which provide the capability of remote start and stop, controlled by the main master machine. As the running time grows longer, both the failures and the power usage in computer systems are thought to be increased.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 65% of the total text.

Intelligent Distributed System

       Described herein is a control method for improving the
reliability of distributed computer systems by shutting down the
power of unnecessary machines and for increasing the availability of
systems by providing the fault tolerance capability.  The technique
is widely applicable to all distributed computer systems which
provide the capability of remote start and stop, controlled by the
main master machine.  As the running time grows longer, both the
failures and the power usage in computer systems are thought to be
increased.

      The method will be best understood with reference to the
flowcharts which represent the controlling algorithm.  Let us suppose
the system is composed of M (M>1) machines and each machine can be
both master and slave machine at the same time.  Also assume that the
capability, as master and slave per unit time (a, b), is accurately
known for all the machines.
      1.  The main master machine calculates and records the current
job amount for the unit period in the system.  Next, it analyzes the
job amount for the next unit period, based on the previous trends and
then knows how many machines are required (Fig. 1).
      2.  The machine defines the order of all machines including
itself, referring to the accumulate powered-on hours of each machine
(Fig. 2).
      3.  Finally the main master machine determines the machine
configuration for the next unit period, i.e., which machines...