Sometimes I wonder if vendors become obsessed by maintaining compatibility with their old products that they are really hamstringing themselves. What is called pedigree could also be termed legacy and in IT, legacy doesn't really carry a positive image.
I wonder if for example 'OnTap 8' would be an established product by now if NetApp had not spent years trying to integrate Spinnaker and OnTap? And simply started afresh?
How much more mature would wide-striping and thin-provisioning be if EMC could have started again and not have to graft it onto existing systems?
And as customers, how much of the blame do we carry? We'll change vendors and hence infrastructure but suggest that we stay with a vendor but change the infrastructure; that'll cause major eruptions. The thought of having to migrate scripts, procedures etc, so we continue to hamstring our incumbent vendors.
Then 'new' concepts such as Cloud come along and still we insist that our incumbents maintain the legacy stuff ad-infinitum. I really think that some of the big vendors should be a bit braver about this but I doubt they will; in some ways this is good because it allows smaller vendors without a legacy to come through and challenge.
Also I suspect that there is a lack of confidence with some vendors; if you are going to force a customer to change infrastructure, processes, procedures and hence incur a cost of change; well they'll look elsewhere won't they? However if you really had confidence in your products, you wouldn't worry about customers looking else because yours is the best, isn't it?
Martin
It’s an interesting conundrum. As vendors bring out new products, some changes are subtle, some not so. Even the small changes have a big impact. Now if customers weren’t so badly organised, change would not be a problem. In fact, coping with a vendor’s change in hardware would be no different to moving to another vendor altogether.
Too little thought is placed on the future and how to remove a technology. Most thinking revolves around how to get the technology into an environment in the first place.
I agree that customers do carry a substantial amount of the blame here. Organisationally, there is a lack of planning for migration and exit. Currently involved in a project for a system which has at least a fifteen year lifetime and at least in the storage arena, we are assuming several technology refreshes. And actually, completely changing the storage technology would not be impossible.