Browse Prior Art Database

Allocate Records in Cache

IP.com Disclosure Number: IPCOM000038785D
Original Publication Date: 1987-Mar-01
Included in the Prior Art Database: 2005-Feb-01
Document File: 1 page(s) / 12K

Publishing Venue

IBM

Related People

Legvold, V: AUTHOR

Abstract

Before the cache can be used to store records, the defective areas of the cache must be found and the required structures to process these records must be allowed in defect-free areas of the cache. This work is performed in two steps. The first step is done during Initial Microprogram Load (IML) based on the size of the cache. The second step is performed based on data supplied by the customer. Defect-free bytes of system storage are allocated to the other Directory Area. This is required because errors detected accessing critical areas of the subsystem storage indicate potential data integrity problems. The record slots are then allocated using weights specified by the host. The number of slots of the first size specified by the first weight will be allocated.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 79% of the total text.

Page 1 of 1

Allocate Records in Cache

Before the cache can be used to store records, the defective areas of the cache must be found and the required structures to process these records must be allowed in defect-free areas of the cache. This work is performed in two steps. The first step is done during Initial Microprogram Load (IML) based on the size of the cache. The second step is performed based on data supplied by the customer. Defect-free bytes of system storage are allocated to the other Directory Area. This is required because errors detected accessing critical areas of the subsystem storage indicate potential data integrity problems. The record slots are then allocated using weights specified by the host. The number of slots of the first size specified by the first weight will be allocated. The number of slots of the second size specified by the second weight will then be allocated. This process continues for the third and fourth sizes. This process repeats when one pass of all four sizes has been performed. If the area to the beginning of the subsystem storage allocated to the Storage Directory was not used for full track buffers, the allocation of the record slots continues with the high half of the subsystem storage allocated to the Storage Directory when the beginning of the subsystem storage assigned to the Storage Directory has been reached. Record slots will be allocated until the next record allocated would overlay the last directory entry. To save time durin...