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

FlashCopy Source Read Minimising Background Copy Algorithm

IP.com Disclosure Number: IPCOM000188465D
Original Publication Date: 2009-Oct-09
Included in the Prior Art Database: 2009-Oct-09
Document File: 1 page(s) / 43K

Publishing Venue

IBM

Abstract

The multiple target FlashCopy as a cascade algorithm described in US patent 7386695 will bound the number of write operations to FlashCopy target disks required for each host write. However, it does nothing to minimise the number of read operations of the source disk required in order to carry out the background copy process for each FlashCopy map. The method for constructing the cascade and the scheme for ordering the background operations almost guarantees that the disk selected for the read operation is the source disk - i.e., the same disk is used as the source for all of the maps in the cascade. This will increase the load on the production disk leading to a possible loss of performance. This disclosure describes an ordering scheme for the background process that will lead to less reads of the production disk. The scheme uses the background copy process of the next map downstream to set the starting point of the background copy.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 58% of the total text.

Page 1 of 1

FlashCopy Source Read Minimising Background Copy Algorithm

The multiple target FlashCopy as a cascade algorithm described in US patent 7386695 will bound the number of write operations to FlashCopy target disks required for each host write. However, it does nothing to minimise the number of source reads required in order to carry out the background copy process for each FlashCopy map.

    The background copy process reads the source data and writes this data to the target disk. The algorithm needs an order in which to perform these reads and writes. Such a scheme may start reading from the beginning or end of the disk for example. The cascaded FlashCopy algorithm means that for each background copy operation, each map will traverse the cascade upstream, looking for data on the target disk of a map. If it doesn't find the data on any target disks then it reads the data from the disk at the head of the cascade.

    The method for constructing the cascade and the scheme for ordering the background operations almost guarantees that the disk selected for the read operation is the source disk - i.e., the same disk is used as the source for all of the maps in the cascade. This will increase the load on the production disk leading to a possible loss of performance.

    This disclosure describes an ordering scheme for the background process that will lead to less reads of the production disk. The scheme uses the background copy process of the...