|
New Year 1998
|
Your computer system is an essential business tool. It was purchased to sit quietly in the corner, managing your data and running your critical applications so that you could concentrate on making your business grow. It’s been doing it for years - too many years. The system on which your business has come to depend has not kept up with the increasing demands of your organisation. You want to add more users, but it can barely support the load it already has. You need more disk space, but upgrades for your model are no longer available. You need to enhance your software, but the company that provided it has gone out of business, or the staff who developed it in-house have left - and took the documentation with them in their heads.
Any of this sound familiar? A lot of it, probably. It is becoming an increasingly common problem in the MultiValue marketplace: the Legacy Effect. Systems that are an integral part of your business, holding vital data or running crucial applications, are now long overdue for their concessionary bus-pass. You need to move, but how? And to what? There are three areas you need to consider:
IDENTIFY THE PROBLEM
No-one knows your business like you do, so if you can clearly identify the problems you are facing, you stand a better chance of obtaining the solution you need, rather than the one the next shiny-faced salesman wants to sell you. There are a number of scenarios that spring to mind:
How much of this do you need to resolve now? This may seem like the ideal time to overhaul your entire IT strategy, taking a ‘Big Bang’ approach, but this can involve a lot of expensive kit, software re-writes, user re-training and the potential for a great many things to go very wrong...all at once. Decide which is your most critical problem and act accordingly.
MAKING THE MOST OF WHAT YOU HAVE
Over the years your company has made huge investments in its IT systems: up
front costs for hardware, peripherals and software, and hidden investments in
the user training, onward development and application dependence.
You would be wise to attempt to preserve as much of this investment as practicable.
With the modern incarnations of the MultiValue model, such as jBASE, mvBASE
and D3, it should be possible to retain and exploit at the very least your software,
and perhaps your peripherals as well. The users see something familiar too,
so the learning curve overall is reduced to a gentle stroll.
Preserving your investment need not mean sacrificing the latest technology.
mvBASE et al provide a gateway to the world of network computing, client/server
and desktop integration that you can phase in at your leisure, as budgets permit
and the business demands.
LOOKING TO THE FUTURE
Where do you want to go next? Undoubtedly the industry trend has been towards client/server, network computing, with all that that entails. If that is your goal, what steps do you need to take to achieve it? What hardware will you need? What are you going to do with your existing applications - go ‘thin client’ and front-end them with something like wIntegrate or Pixel’s HostAccess, or opt for a complete rewrite? What are the costs of each step: in time, in money, in training? If you are considering more than one possible platform, can the vendor perform a test portation so you can get a feel for it yourself?
This is linked inextricably to your current situation: the solution to the existing problem will necessarily be the first step on the ladder towards your new goal. It is neither practicable nor desirable to set out on the journey without knowing the destination.
Consider an example. Company ZZZ Limited had an old machine that was expensive to maintain. It was groaning under the weight of users spread across head office and a number of remote sites connected via leased lines. They needed to move to a more powerful platform but had a number of constraints:
They opted for a MultiValue solution based on Windows NT. A test portation showed that their existing applications could be migrated with little change required to the code. All their existing peripheral devices such as modems and printers could be utilised, and the new environment supported dumb terminals and PC workstations with equal facility. Transition to the new system was accomplished in around 24 hours at a weekend, with zero downtime. Total cost: less than three years’ annual maintenance on their old system.
Already they are reaping the benefits. The system load is up by 20% with scope for 512 users, performance has improved by at least an order of magnitude, and to the average user the system looks no different than it did before. With the NT-based system, they have removed their hardware bottleneck without disrupting daily business, and now have the potential to develop a phased client/server strategy at their own pace.
Tony Coutts
SOS Limited
Last Updated: 09 November 1998