Big Blue's storage software duo

It's not the usual procedure at IBM, but when it comes to data storage software, Laura Sanders and Bruce Hillsberg share responsibility for running the show.

Sanders is vice president of storage management products in the company's Tivoli software unit, which makes products for monitoring and managing storage equipment. On the hardware side, Hillsberg directs storage software strategy and technology in Big Blue's systems group.

 IBM sees storage as a proving ground for utility computing.

Storage software has become increasingly important to organizations eager to increase the utilization rate of their gear and trim personnel costs through automation. What's more, the storage industry has been talking up the concept of information lifecycle management, or ILM. That involves software that moves data from one type of storage device to another, as it becomes less valuable or needs to be permanently archived to meet a legal requirement.

The Tivoli brand stems from IBM's acquisition of Tivoli Systems in 1996. Meanwhile, the hardware side has been working on its own software technology. Sanders and Hillsberg recently discussed IBM's take on information lifecycle management and what they believe is shaping demand for storage.

Q: You have not made a big deal out of information lifecycle management. Is that conscious, or are you planning to do something along those lines?
Hillsberg: From our perspective, we have been doing this for decades on the mainframe. HSM (hierarchical storage management, which refers to putting data on different classes of storage gear) has been part of the ZOS offering for some time now.

We have been providing this for some time for our customers, both in the mainframe and in the open systems environment. Is it important? Yes, but it is a piece of the solution. What we are trying to do is help customers create more of this on-demand view of the world. Making sure that your data is in the right place is one aspect of it, but there are other aspects equally as important--things like improving application availability.

So when I go mess around with that storage, or something comes off-lease, I don't have to take down my mission-critical applications to do that. It's also key to implement the intelligence in the system so that, if I am managing a thousand servers and all that associated storage, I don't have to have people worrying about Bruce creating a file on the right piece of storage.

We continue to invest heavily in TSM (Tivoli Storage Manager). We have made some recent changes to make it work well with some of these compliance regulations, but we already have laid the base.

Sanders: EMC rushed out to buy Documentum. We have Content Manager already, which does a broader set of functions than what EMC purchased, and TSM.

We actually use the TSM code as the data store under there, because

The No. 1 thing that takes you down is human intervention, and that is why everybody is into this automation stuff. --Laura Sanders

it knows where to put things. Let's say you're CNN, using Content Manager, and you pick your movie, and you say, "OK, I'm never going to use this clip; put it over here." And we just place it for him. And then if you want to go back and say, "Oh, gee. That was a movie of Gates. We might get subpoenaed for that; I never want to erase it." You can do that. And you can do that with your e-mail, with your pictures. So we are kind of enjoying the whole ILM craze, because everybody is going, "I got to get this, I got to get that." We are going, "Mmm-hmm."

Hillsberg: It is sort of cool. EMC is doing a lot of demand generation for us.

Q: So, you are just selling this stuff without even having to talk about ILM?
Hillsberg: Customers ask that we talk about the capabilities of HSM and archive. What is exciting for a lot of our customers is that a lot of them have TSM, and they did not realize that they have these capabilities. They already have it, and they can now exploit it further.

Sanders: We are getting more requests for ILM-type briefings, but we do walk in and talk about the whole on-demand concept.

Q: What's driving demand for storage? Is it things like compliance?
Sanders: I think that the first thing driving demand is application uptime. The No. 1 thing that takes you down is human intervention, and that is why everybody is into this automation stuff.

But just like you can only get so much out of the hardware, and then you go to the software, you can only get so much out of automation, until you simplify the layer below it, and that is why virtualization is important.

The next thing after that is "OK, I'm up. What is my group's productivity?" You do not want your people hanging out and making sure you are backing up or making sure that the system's running. You want them working on the next project. And then you get to optimization. If we can provision on demand, you can have a much less large infrastructure, still as robust, and migrate it to where you need it, when you need it. That is the whole idea of on-demand.

		Get Up to Speed on... Utility computing Get the latest headlines and company-specific news in our expanded GUTS section.

Hillsberg: If you look at the trend, storage needs are increasing. That is a curve that just keeps going up. There are a lot of drivers today. There are new applications coming online. Data and information are competitive assets. The regulations certainly drive more storage utilization. So, as that goes up, you also see that storage and storage management are becoming an increasingly large percentage of an overall information technology budget.

In the past, when people needed more storage, they'd buy more storage. They keep buying more storage, and then that just keeps driving these curves up, and people cannot afford to keep hiring more people. They need new tools that are going to help them start to level off some of those expenditures and the human intensity that is required to manage that.

What we are trying to do is really simple. We are trying to provide you an infrastructure and a set of management tools that will allow you to get off of those curves. We are trying to do that in a way that accommodates not only IBM stuff but the stuff you may already have today.