There are two values that CA Disk can calculate for a cluster, depending on whether or not the cluster contains any spanned records. (The cluster must actually have at least one spanned record, not merely be defined with the SPANNED attribute.) If the cluster has no spanned records, the only value to be calculated will be the number of non-spanned records that can be added. This is based on both free space in used control intervals and also on all unused control intervals. For control intervals that already contain data records, the calculation is based on the free space in the control interval and the average record size of records processed to that point.
For unused control intervals, the value is based on how well records fit into the used control intervals.
When a cluster contains spanned records, the calculation becomes more difficult. In effect, CA Disk tries to simulate the addition of both spanned and non-spanned records at the same percentage as that which occurs in the data itself. Without going into the elaborate details, CA Disk basically does the following:
Because of the method used in simulation, the percentage of new record additions can vary slightly from the actual percentage found in the data.
|
Copyright © 2015 CA Technologies.
All rights reserved.
|
|