Browse Prior Art Database

Data Migration with Volume Merge

IP.com Disclosure Number: IPCOM000222178D
Publication Date: 2012-Sep-24
Document File: 2 page(s) / 32K

Publishing Venue

The IP.com Prior Art Database

Abstract

This publication introduces a method and apparatus to perform a data migration as high priority I/O without impacting system performance.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

Data Migration with Volume Merge

In storage controllers data migration from one storage controller to another can be performed as a background process with low I/O priority which can result in a very long time to complete the migration - at times taking many days especially when many volumes from the source system are being migrated to a new target system. Even if the migration is tagged as high priority I/O it can affect overall system performance because the migration is affecting volumes that are also receiving host I/O.

This publication describes a method to perform the data migration from an external storage system to a data migration volume that is inaccessible from a host system. This data migration volume is used to hold all the data from an external storage system volume that is being migrated.

Initially the host volumes and the data migration volumes on the target storage system will be created the same size as the source volumes that contain the data to be migrated using a new volume creation module which will automatically create the data migration volumes.

This includes a new feature in the user interface (ie graphical user interface or command line interface) that allows selection of offline data migration which will create the data migration volumes automatically when the host volumes for migration are created on the target system.

The bitmap of the volume to be migrated (also referred to as the original volume on the source system) is retrieved once the data migration is defined and built on the host and data migration volumes on the target system. The server is then mapped to the host volume on the target system and switched over to the target system so that all new I/O is written to the host volume only (not the data migration volume). All read requests will initially be directed to the original volume on the source system. As new data is being written, any read requests to that data will be directed to the host volume. In addition as the data migration volume is getting the data transferred, read requests have the potential of being read locally from the data migration volume which also improves read performance by not having to read from the original volume which could be located in a remote site. In current art, all read requests for data not on the host volume are sent to the original volume which can result in lower performance in reads as...