Browse Prior Art Database

Internet Data Center Value-Add Feature: Total Performance Capacity Engineering Concept Disclosure Number: IPCOM000015215D
Original Publication Date: 2001-Nov-01
Included in the Prior Art Database: 2003-Jun-20

Publishing Venue



This invention disclosure describes a concept allowing one to only engineer critical core components (e.g., firewalls, web application servers, routers, etc.) of any infrastructure architecture to a 50% maximum load at any given point-in-time. This disclosure, describing the Total Capacity Performance Engineering concept, implies that any critical unit is virtually 100% consumed (regarding machine resources capacities) when it is operating at 50% capacity levels. This is necessary in order to allow for the additional buffer required in failing situations of other components that is, when they fail and require transfer of processing capabilities to other devices. The problem is that highly utilized components fail from time-to-time, thereby needing another component to re-direct the entire load for fail-over processing. This disclosure, the Total Performance Capacity Engineering Concept, solves this problem of insuring that 100% processing capability for an integral fail-over state exists, from one component to another component. The advantage to using this fail-over concept is that by planning and maintaining 50% of any component's utilization as "total capacity engineering," then transferring this 50% capacity to another machine also maintaining 50% capacity allows for 100% full machine utilization in an unpredictable outage situation, i.e. 100% high-availability. This concept is a pre-design concept, not one which is easily implemented post-design (or, post-implementation). Traditional infrastructures can overlook this important design concept, often times realizing this as a design quality point too late in the life-cycle, or maybe sometimes not at all until failure transfers are attempted. This design concept will insure that adequate fail-over capacities exist in each of the targeted components, if ever required for recovery, should an unplanned outage unexpectedly occur. By factoring this concept of Total Capacity Performance Engineering in all infrastructure design activities, the overall design quality of the targeted environment(s) will be insured of maintaining high availability. This concept will prevent unplanned peak surges to other components, as total loads are transferred from other failing devices.