Microsoft's Muglia talks Windows Azure

By Ina Fried

A long time ago, Bob Muglia worked on a Microsoft project designed to offer a variety of services in the cloud. That effort, known as Hailstorm, didn't exactly go gangbusters, and Muglia's career took a detour.


But both Muglia and Hailstorm are back. Last month, Microsoft elevated Muglia to divisional president, a recognition of the success he has enjoyed as head of Microsoft's server software business.

As for Hailstorm, the name is gone, but many of the concepts are back, as part of the Windows Azure platform that Microsoft announced in October.

In December, Muglia spoke to ZDNet Asia's sister site CNET News about Windows Azure, the cloud in general, as well as the economy. Here are some edited excepts:

Q: Is this supposed to be a slow-motion rollout with Azure?
Muglia: The way I sort of describe it is, it'll be phased--there's a whole broad set of services. You'll see some of those services go to production next year; exactly what and when, we're still working through. People are able to begin to develop right now, of course, on it, but it will happen over a period of time.
And the other thing right now is, people are still very much kicking the tires. We have quite a bit of tire kicking going on, a lot of people provisioned on the services right now, and so far, things have been going well.

What are the kinds of things that you think people will want to run in Azure?
In terms of the classes of applications, I think you'll see two initial ones, though it's fair to say that people may have an interest in running in this environment any application they would want to run on-premises. But the initial ones I think would be your Web-style applications, which tend be Internet-connected and need geodistribution.

The other class that I think is really interesting is anything that involves working in partnership with others: supply chain sorts of applications, business-to-business, Electronic Data Interchange, those sorts of classes of applications in which you need to connect multiple organizations, and you need to deal with authentication, and you need to deal with network connectivity.

Today it's very complex with virtual private networks and password management, and a whole nasty set of problems to deal with, and Azure has some built-in services to simplify those things, including a service bus to go through firewalls and connect things over the Internet--again, any system to allow people to authenticate. Those are basics that are fundamental, that everyone will really need in this kind of environment.

So, I think you'll see those sorts of things emerge initially, but then you could just imagine all sorts of things. You could imagine people using it for (high-performance computing) applications. That's an area we're looking at, and we certainly are having conversations with a number of academic and other organizations.

In terms of the movement toward the Microsoft-hosted versions of its server products, are there some interesting things that you have come across, as you've had to do the work to get ready for that?

There's a ton. I mean, one is, you need to move so that everything works across the Internet, which is just the right thing to do, anyway. Another thing you see is the need to have what we call multitenancy. So, to get scale on these things, you can't be dedicating even a virtual machine to a company. You need to be able to support many users, many organizations within a single instance of an application, and so that's an attribute.
There's a whole set of really interesting regulatory things that you hit when you go across the world attached to this. Turns out, some of our products, like unified communications, have VoIP (voice over Internet Protocol) capabilities. Well, you go and take that to many countries of the world, and the call those telephone companies. They call you a telco, and all of a sudden, there's a conversation about being regulated like a telco, you know, in some countries around the world. That may or not be pleasant.

There are issues about data and where data can reside and not reside, and so that's why when you spread geographically around the world, there's a wide variety of new issues that open up that are quite, you know, quite interesting--billing issues, because obviously, there are different issues with the way the banking systems in different countries work.

Let's talk about the economy. What are you seeing when you talk with customers?

People are afraid. I mean, I think we're all a bit afraid, at some fundamental level, because we don't know--no one knows where this is going to land in the long run.

No one really is clear as to how far the contraction is going to go and how long it's going to happen, and then there's a lack of clarity also as to how we get through this. Are we going to be tight six months of the year, then boom!? Well, maybe. That would be kind of a good viewpoint of things right now. Or is it going to be a longer period of time, with a medium period of time with sort of a slow growth.

I sort of always come back to a belief that the fundamentals will drive all of these things, and ultimately, it means that people have to produce things that others value that helps to drive the overall society forward and, you know, generate something that is of sustainable long-term value. Ultimately, one of the key things is, how can we make companies and individuals more productive and able to work together better?

So, I guess I have two questions somewhat related to that. One is, how about for you? As a business manager, obviously, you manage a fairly large business. What are the things that you might have done, had the economy continued, that you're not doing now? What are some of the things that remain priorities, and what are some of the things that you're going to let happen slower?
Well, certainly, there's no question that Microsoft's not immune to the circumstances. There's no question about that. So we have slowed our growth. We are still growing as a company, and (the server and tools business) will grow overall this year, though I admit that we did most of our growth in Q1.

We actually were incredibly successful in bringing a lot of folks on in Q1, so we would have had to slow (hiring) under any circumstances because we're outachieving our plan, but we've slowed considerably.
So if you take some of the areas in the database space, like some of these areas around business intelligence and data analysis, we're actually investing in some of those areas. But we're taking resources off some things that don't have the same kind of results and long-term potential for us to have returns, one of which was pretty public recently: OneCare, where we, you know, decided to refocus that effort into a much more narrowly focused free antimalware offering instead of providing a broader suite.

Are there other things about which you, as a business leader, are saying, this is going to have to wait a little or move slower?
Yeah. I mean, there certainly are. I think that we've looked inside, at what we're doing really in almost every one of our groups. If you look at almost every one of the things that we're doing and say, OK, there's a set of things we want to do in management, let's tighten the belt a little bit, as to where we're going. Yet we're continuing to invest in this whole virtualization and management space, coherence with Azure, all those sorts of things we're continuing to invest in.

So in each one of our business areas, we've looked at how we could reallocate and refocus, and then across the board, we've made some fairly fundamental shifts like we did with OneCare.