Assessing the impact of continuous change on the storage industry.
On December 6, 2001, executives from Colorado-based data storage companies met at StorageTek's headquarters in Louisville, Colorado. Mark Ferelli, Editor-in-Chief for Computer Technology Review, hosted art illuminating roundtable.
Mark Ferelli: I'd like to make everybody here a futurist. We've established that by 2004, IT spending will be two dollars for every three on mass storage. Are we talking about connectivity, are we talking software, are we talking hardware, are we talking all of the above, and in what kind of mixes?
Juan Rodriguez: It's all of the above. Storage is becoming cheaper, productivity is becoming cheaper, and software never ceases to have problems with it. The issue of managing the whole system becomes much more problematic. I think the issue of management continues to be probably the biggest consumer of that cash, the cost, and management of data storage. The issue of interoperability requiring a lot of manual intervention or human intervention.
Jeff Laughlin: I believe that it certainly is going to be a combination of all of the above. I believe out of that sixty-seven dollars out of a hundred dollar bill, probably half of that is going to be spent on software, and I mean microcode, firmware, and post resident software as well as fabric resident, which happens to run on a processor hidden inside the fabric. Kind of taking an administrator's view of things, because they are the people that have to put up with all the stuff we have propagated onto them. I envision a world where software is helping them with this whole notion of availability. Things are just not going to be allowed to go down for any length of time.
Gary Wright: We have seen that eighty percent of the storage is purchased by twenty percent of the companies in the world, and looking at those larger companies, what we see is the future, a global storage network that offers automated policy-based life-cycle data management. And if you look into that crystal ball I really see four areas where a customer will be spending the money in the IT budget. A lot of it is going to come from the infrastructure so there will continue to be hardware expenses that are spread out over lots of different devices in the connectivity. Second I think is the long-distance networking charges. The third is intelligent software that is going to provide the automated intelligent management. And the fourth area is services, to pull all this together. These are large-scale projects that the typical customer is not going to be able to do in house, and it will require some pretty sophisticated services to make this work.
Michael Harrison: 2004 may seem like it is a long ways away, but it's not. I believe there are shifts in the industry that are happening today that are going to take longer than 2004 to take hold. One of the main ones I think is in the area of storage resource management. And in the majority that is going to be a software charge that customers will need to bear. In the 2004 time frame we will be well on our way to changing the model to be much more heavily weighted on the software side, devoted to the management of storage rather than the hardware.
Steve Lambourne: I don't think people costs are going away any time soon, and I think that a percentage of their budget will go down and be replaced by software and services. I think that we here as well as many, many organizations are looking to outsource stuff that they no longer want to do. They want to get it out of their organization. That's a whole other topic, why would you want to relegate storage management when you say that's so important to you and data is an asset, why would you want to relegate that to somebody else? But that's a whole other issue.
Mark Ferelli: Then the death of storage access providers has been greatly exaggerated.
Robert Sellinger: I think the trend shifts are going to be a little more subtle than dramatic. I think that manpower and labor content is going to continue to grow gradually but I would also include services within that. Software, certainly content is going to go up, and I think the charges from remote bandwidth for data traffic, for the offsite recovery et cetera is probably the biggest delta I would expect in the typical IT budget. In terms of all those things as a percentage, what is going up and what is going down? I think all the devices are going down, all the amounts on disk or tape, I think it is probably true even on things like controller switches, we certainly see it already in terms of things like servers. In general I think all of the hardware components, while the unit volume may increase, cost per component is going to take that overall smaller percentage of the total.
Dick Search: I have to agree with both Robert and Mike, 2004 is tomorrow. Everybody believes that services, resources, and software are going to be a much bigger part. But it won't be substantially that much bigger by the time we get to 2004. That is going to be a longer evolution. Although the cost of hardware is going down, both networking and physical storage demands are still skyrocketing. So it's still going to be a third of the overall spending because where you used to buy a terabyte, now you are going to buy a petabyte. So it's probably going to be a third, a third, and a third. The hardware side of it includes both the network infrastructure plus the storage.
Mark Ferelli: As part of your prognostication I would like to ask about the installations of network attached storage and storage area networking devices. Will we get past the hesitation that SAN has seen?
Dick Search: It's going to take a little while. There is going to be a convergence of NAS and SAN over time. In fact IDC is actually going to stop counting them as separate things because it's network available storage somewhere. The key is just what we talked about very early on, interoperability and management. As soon as people get to the comfort level that all that complex stuff that's in the cloud works really well, then it is going to continue to grow, but it's going to take a while for that to happen.
Joe Raschke: The ratio of management costs in the human factor of managing it is still going to be there significantly. The disk grows. You're managing terabytes or petabytes or in our case we do yottabytes of storage. At that point you still have this large management cost that is sitting on the top of it, so you really grow into that. And you are also looking at storage not from a disk, and not from a tape perspective, it's from a network perspective.
Walt Hinton: There is a fair percentage of every enterprise budget that is spent on something that I believe is arcane. And that is the trucks that come around every night, and accept your tapes, and take them off to Iron Mountain. It's a billion-dollar business and I predict by 2004, it will be dead. Having people come around with trucks to take your tapes away and put them in a vault where then at best you can get your data back in ninety minutes or most likely the next day when the trucks come around again is just lousy. With the cost of band-width at forty-five megabyte circuits being fifteen hundred dollars a month, there is no reason why a remote electronic data vault can't displace Iron Mountain.
Meg Heller: I have to agree to a large extent with everything that's been said, storage resource management, real critical software, and how you can lower your total cost of ownership and your resources and what they are focused on is going to be real critical. SLAs are becoming more prevalent, customers want them down to the application level, and software management's really going to help us do that. Hardware infrastructure, that's going to be there, a third to the software, a third to the hardware, and then a third to services, because people are helping us design, architect, and implement their solution because they don't have the manpower and skill set to really do that and they want us to help manage that.
Bill Reed: The layer that we are calling storage management today as a separate application has to sink down in a hierarchy to almost a middle-ware functionality in a computing hierarchy, so that it disappears. One of the most consistent things we hear from customers is, "We would really rather not have to deal with this." There is an almost unanimity in that. You don't get promoted because you did storage better; you get promoted because you implemented the ERP system better than expected. So, to the extent we are able to invest in making storage disappear as a management issue, that's where we are going to be able to get a premium dollar from the customer and that's what they are going to pay for it, so I see that growing very fast.
Mark Ferelli: Bill brings |up a really important point, which is the fact that applications are king in this business, and will never condescend to any secondary role. And anything storage does to make an application in the enterprise easier, quicker, better, that's where dollars are going to be spent.
Juan Rodriguez: Let me say one thing. I probably became aware of how fast data was growing twenty-five years ago, when I heard that it had been doubling every year, and it has been doubling every year ever since. I think that is one of the things that we deal with, and quite possibly this proportion of cost between hardware, software, and service has remained relatively constant. The issues of managing the data I think are geometric. And it's only the software that can allow us to manage that. There is a certain balance of forces here where you can't go any faster because the people we are dealing with, the people who are managing the processes, can only grow so quickly in their experience, and they are limited.
Mark Ferelli: One last thought, Meg, something you need to take back to Scott [McNealy, Sun's co-founder]. The Sun mantra at one time was that the network was the computer. 1 would suggest nowadays, considering spending, the network is the storage, and trying to avoid that particular inevitability is time badly spent.
Meg Heller: That is our mantra for the storage division.
Mark Ferelli: Go tell Scott. He needs to buy in on the proposition. Thank you very much everybody and we will see you in the papers. Take care.
Jeff Laughlin, Director of Strategy, Automated Tape Solutions, StorageTek
Robert Sellinger, Chief Technology Officer, Chaparral
Joe Raschke, Business Development Manager, Yotta Yotta
Richard Search, Vice President, Marketing, McDATA
Gary Wright, Director of Marketing, Enterprise Storage Group, Compaq
Steve Lambourne, VP Product Marketing, Hitachi Data Systems
Michael Harrison, Director of Business Alliances, IBM
Walt Hinton, Chief Architect & Senior VP of Design Services, Managed Storage International
Juan Rodriguez, Chief Technologist, Exabyte
Bill Reed, VP of Marketing and Business Development, SpectraLogic
Meg Heller, Director, Storage Services, Sun Microsystems
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Enterprise Networking|
|Publication:||Computer Technology Review|
|Date:||Mar 1, 2002|
|Previous Article:||Advancing New Enterprise storage beyond the campus.|
|Next Article:||Ultra-Wideband technology to offer new opportunities for wireless video, networking, Part 2.|