Written: June 2001 by Geoff Norman of Scrittore
Reproduced with permission. This article describes how storage management should be moving forward to a seamless, self healing architecture.
Introduction One notable feature of recent enterprise storage
evolution has been a consistent emphasis on matching computing activity to business requirements; IS has
moved from being an analogue of business activity to forming an integral part of it, or even to actually being that activity. New business models - exemplified by B2B e-business and thus much more fundamental than selling via the web - call for new methods of managing systems, and there were frequent references to close co-operation across enterprise and, hence, technical boundaries. scrittore papers have been stressing the importance of these changes for several years, and suggesting ways in which IS has to respond.
One of the most demanding challenges, which is receiving a great deal of attention from the vendors, is to manage the huge - and rapidly increasing - volumes of data that are involved in commercial computing. This paper is intended to suggest new approaches to data management as it moves towards proactive rather than reactive operation. The method used is to look at several examples of systems management at work; the idea is that each should be technically feasible with no operator intervention.
The series of examples The first example looks at the level of automated
management that could be achieved today; the others examine the future, with the last two showing how crossing enterprise boundaries will add to the demands made of storage management. They all expand on the data management themes that have been introduced in scrittore research over the last couple of years, particularly on the topics of data recovery and recovery avoidance.
The implied technicalities of delivering future enhancements are not simple, and timescales for
availability are lengthy - depressingly so in view of the development dollars and IQs devoted to this area, and the predictability of enterprise demand for the levels of function outlined.
It is axiomatic that business continuance prefers 100% availability and demands the swiftest possible recovery from any problems. It is also worth noting that in all cases the interpolation of a Storage Area Network (SAN) between applications processors and storage devices makes life a lot simpler and management a lot more flexible. This can be taken for granted but will be referred to where it makes a particular difference.
scrittore
See the next page, or use the left hand frame to navigate between the three timeframes.