In-Depth

IBM Touts zSeries to Simplify IT Infrastructure

Big Blue discusses the idea of reducing network and application tiers, and offers a model that combines zSeries technology with Intel-based server blades.

• By Stephen Swoyer
• 09/23/2003

Last week, IBM’s Systems Group hosted an “Infrastructure Simplification” day at its Somers, NY campus. The event was largely given over to a discussion of an idea—infrastructure simplification—in which organizations, having consolidated from many to few servers, should also think about reducing the network and application tiers in their environments. One surprising upshot of the day was a model offered by Big Blue: a combination of zSeries mainframe technology and Intel-based server blades.

We spoke with Pete McCaffrey, program director for zSeries marketing, about zSeries’ role in IBM’s infrastructure simplification.

What does IBM mean by infrastructure simplification?

We position it as the next wave, after the traditional server consolidation that’s been occurring over the last couple of years. What we highlighted are sort of the key enabling technologies that are allowing customers to sort of further reduce or collapse their infrastructure to fewer tiers of servers, fewer tiers of networking, storage, etc., then ultimately tie these together by greater overall value. That value would show up in the way of reduced cost, flexibility, provisioning of resources, so that customers could react quicker to change, so that there’s a number of motivating factors.

Today [last Tuesday] you touted a combination of zSeries mainframe and server blade technologies as a textbook case of what you’re calling infrastructure simplification. That doesn’t seem to be the most obvious of pairings.

That’s right. There’s sort of the counterintuitive notion of mainframe+1, the mainframe sort of representing the ultimate in scale up, whereas blade centers are sort of the ultimate in sort of a scale-out environment. We’re seeing customers getting the two centers of gravity and in some cases tying them together to match the application to the right technology.

Is Linux the unifying technology here?

It’s one of them. The trend we’re seeing is customers sort of consolidating a number of these different elements [distributed applications or services] to one or the other technology [mainframes or server blades], with Linux running everywhere and with virtualization technology on the mainframe. So we have some customers that are focused on consolidation on the mainframe using Linux, others focused on consolidation into a blade-type of environment … customers do both. In some cases they may move elements to the blades, in others, they’ll move it to the mainframe environment.

Is infrastructure simplification a deliverable or a service?

It’s really more of a trend that we’re seeing. If you think about it, go back a couple of years and even today, customers were in situations where they had literally hundreds of servers or in some cases thousands of servers, they were going through the server sprawl. They went through this period of server consolidation, they got bigger servers and leveraged partitioning and virtualization technologies, and reduced the number of servers that they had.

Simplification is really kind of taking it to the next level. I’ve got the number of servers down, but let’s reduce the number of tiers. Today I’ve got three or four tiers. I’ve got security and caching servers on the front end and Intel servers at the edge. Infrastructure simplification says let’s further reduce that environment.

Is infrastructure simplification something that you’re positioning for customers with existing investments in mainframe or server blade technologies, then? Or are you saying that customers should invest in these technologies to further consolidate away from their existing platforms and tiers?

What we are doing is highlighting the idea and the benefits. In our examples, we focused on blades and mainframes, but what we’re saying is that there are other scale-up technologies that they could use—it could be a zSeries mainframe, it could be an iSeries mainframe, it could be the Unix-based high-end scale-up server. Similarly, a customer could leverage blades for scale out or Intel-based servers for scale-out. The technologies that help tie these environments together are Linux, our grid [offerings] and [Tivoli] Intelligent Orchestration software.

In some cases, [infrastructure simplification] customers can accomplish this on a single platform. We’ve got big customers that are running z/OS-based applications, and then in a different partition they may be running Linux-based applications. They made the decision to consolidate across two tiers on a single mainframe.

Okay, if it’s not a deliverable or a sellable service, is infrastructure simplification something that customers are under-taking on their own, or will IBM provide services and other support to enable it?

We do provide support. We have more than 20 Design Centers around the world that work on these kinds of products. We’re there to help customers do this, and we also provide some of the key enabling technologies, like our grid offerings and Tivoli.