Listen to my interview with Dan Foody, vice president of Actional Products, a division of Progress Software, an important infrastructure software vendor.
In this podcast, hear why a stronger focus on business value is essential to successfully manage enterprise mashups, and learn how testing processes have to change to avoid unexpected hitches when mashing up cloud services.
Listen to or download the 8:40 minute podcast below:
---Transcript---
PW: Dan, you've said that, as enterprises start mashing up cloud services, something that you've noticed is that they need to start thinking more about focusing on the transactional layer and how they manage that, rather than the infrastructure layer that IT has traditionally focused on.
DF: Well, yeah. So one of the things that you see is that IT managers and people who look at IT operations, they typically have focused on the 'Is it up and running?' aspect of applications. And as you move into the cloud, the nice thing is, you don't actually have to worry about that anymore, that's someone else's job to worry about 'Is it up and running?'
But even beyond whether something is up and running, an organization still needs to worry about 'Is it actually delivering value to the business?' Is the business being successful on top of those IT systems, whether or not they're running? And if your problem is you can barely keep your systems up, then you've got to worry about that first. But once you're dealing with stable infrastructure, then you get to the level of, now we need to worry about, is the business up and running?
You know, Dan, that's interesting, because one of the things I think that people worry about cloud services and SaaS is, does that mean that as an IT guy I don't have any work anymore? And actually, what you're saying is that, well, yes, you do have work but, in fact, it's more strategic than the firefighting that traditionally you may have been doing in the past.
Yes, it's far more valuable work to be able to associate yourself with true business value. And so a great example from one of our customers was one that — they had a case where every — their IT systems were all up and running and yet they found out — after the fact — that their single most important customer wasn't getting updates for a month, which was a huge business issue for them. Now, if the IT team had been able at that time to focus not on the infrastructure side of, is it up or is it down, but on the business side of, are we delivering the value to our most important customers, they would have been seen as a partner to the business not a barrier in that case.
Yeah, that's a very good example of exactly why this is probably a good move for IT to be looking at — getting their head into the clouds, almost — because they can be more concerned then with the real business needs of the organization they work for.
It's the nature of wanting to go to a utility. You do that because other people take the basic work off you so you can focus on what matters to you.
So I was interested — we were talking about some — a new Progress product, a couple of weeks ago, which is — is it Actional Diagnostics?
Yes. Yeah.
And there were some new tricks in there. One of the things I thought was interesting was this thing called Application X-Ray. And why I find it interesting was because you've introduced this so that people can have a better grasp of what's happening when they do mash ups of different services because there's a need to understand the knock-on effect of services. So tell me a little bit more about that.
Sure. So the general challenge that people have faced when they start to mashup services is that they really only see a very little bit of what's going on. So a lot of the challenges and issues that you face when you start to mashup are that you get mismatches in expectations between how you thought you were using this service and how the service — and how the person or the people who built the service expected it to be used.
And those often manifest themselves in very subtle ways. And so what you end up seeing is that, while you may even be getting a right result from sending in something to your mashup, it might be doing the wrong thing and you don't even realize it. Because you don't really understand what's going on under the covers in the overall mashup service.
So with Application X-Ray, what we're doing is giving people that view into what's really going on inside that mashup. So not only do you see that, if you send in a request or a message or type something into the front end of it and get a result back, not only do you see the immediate result, you see everything that went into producing that result.
So you see all of the activity on the mashed up services. You see exactly how your mashup is using these other services to produce its overall result. So you get really this global visibility — this x-ray into what's going on inside your mashup — so that you can see how every component contributes to the overall result. And that's something that people have typically not had before.
Right, and the reason that's important is because when you're doing mashups, you're actually dealing with services that you don't control and you don't own. And therefore, you need to have a better understanding of what's going in and coming out of those services simply because you don't have the blueprints to hand. Or at least, you might have the service definition that the provider has given you, but it might not cover everything that you need to know. You can't actually see inside the service because the nature of a service in this kind of service-oriented environment is that it's a black-box service that you just have a service contract with — but you still need to have a handle on how you're interfacing with it and what's coming back.
Right. Because, fundamentally, you don't really know how it's going to behave in all the different situations that you're going to put it into. That's just the nature of something built by a third party that you don't control. You won't know exactly how it's going to respond to every situation that you want to take advantage of them in.
Right. And, obviously, knowing that, if you can test for that, then that gives you a great deal more certainty as to what is going to happen when you go live, which has got to be a good thing.
Yeah. Well, I think there's really two aspects to it. That's one, that you want to make sure that you cover all the different cases and really understand how it behaves under all the different circumstances. But there may be cases that you miss as well and so as you go into production, you want to be able to deal with problem resolution as well. So you want to take that same thing that you needed in the development phase to understand what's going on, and bring that over to production at the same time.
You know Dan, the thing that really strikes me when we talk about these things is that it just reminds me of all the stuff that we used to talk about in service-oriented architecture and web services when that started up five, six, seven years ago. And people — I've been reading SOA is dead but it sounds to me as though it's just morphed into the cloud. Do you get that feeling?
Well, yes. In many ways, I totally agree because, when we started off in the whole web services space, whatever it was, eight years ago, ten years ago now, that really was focused around different people hooking together pieces built by others and addressing all the challenges around that. And so, if we call that service orientation, that is absolutely at the core of the cloud today. So we really have morphed into the cloud. There was this offshoot at one point that people started to call SOA which was — which you might refer to as big SOA, which is enterprise architects wanting to get their grubby little hands on everything going on within the organization and take control of everything.
Yeah. Yeah, I remember that.
And they direct the agenda of the organization. And I think that morphed into being what people thought SOA was. And for better or for worse that has largely tailed off now. And so the important parts of SOA, the parts that are adding significant value, and agility, and flexibility for organizations have absolutely morphed into the cloud world.
Leave a comment