I have arrived back in London after spending an enjoyable weekend in Denver. It has been a while since we were able to spend time at home and we took time out to entertain a number of our friends from IBM. Over a glass of wine we were able to cover a couple of topics I routinely return to in these blogs – issues to do with the data center. No surprises this time – the picture I have included here is of a few of us gathered in our bar area.
The data center has always interested me – how it’s put together, how operations are set up; the division of responsibilities between application, system and network management. I have also been interested in learning about the views of those close to the systems and the steps they take to ensure uninterrupted operations. And coming from a mixed IBM and Tandem background, I’ve seen just about every combination imaginable.
I have been to the data center of one well-known U.S. retailer, where the only time they paid any attention to the Tandem was when there was a lot of noise coming from the console printer – loud enough to warrant someone going over to take a look. A few years back I was in the data center of a well-known Australian building society where the only thing the data center manager could tell me was how many MIPs they now had on the floor, and how disappointed he was that their new models no longer had any console lights. How could he convince his executives about the capabilities of the new system if they couldn’t tell if anything was running?
But whether we talked about IBM mainframes or Tandem OLTP systems, what is common across most data centers today is the variety of technology deployed. The single-vendor shop is long gone, relegated to being another relic from the ‘70s. Once client-server computing took hold, in no time we were seeing a very diverse mix of equipment being rolled in to support our business applications. Decisions were being made about the types of hardware purchased based on their suitability for supporting an application, as we slowly edged away from our former position of simply buying applications based on our installed hardware architecture.
As we moved into the ‘90s we really had mastered the integration of platforms, storage and communications. Not unlike cabling a modern, high-end, component hi-fi system, we developed the skills where we could plug and play with pretty much any piece of equipment our company brought into the data center. We were comfortable connecting cables to interfaces in much the same way we plugged any CD or DVD device into a home stereo.
But hidden behind the scenes were still a lot of applications running in complete isolation. We had figured out how to connect all the hardware, but getting the applications seamlessly integrated was still something we weren’t as comfortable doing. Yes, we had piloted some object technologies and deployed object brokers. Yes, we had standardized on system and network management protocols and could consistently interface with running applications with some consistency. But it was the exception, rather than the rule, to find infrastructure in place that would support a diverse mix of applications and that allowed data to unambiguously flow between the applications that required access.
It was with this in mind that I just read a Computerworld article “8 Blazing Hot Technologies for 2008” where I saw that hot technology number 3 was Integration! According to the Computerworld columnist “IT departments have been building islands of applications and data for years, and despite repeated attempts to integrate them, many of those islands persist. It's time to get serious and build bridges, so users can get all the information they need in one place at one time.” Jonathan Eunice, an analyst at Illuminata Inc., a Nashua, N.H.-based IT consultancy, was reported to have said “If you're a database administrator, an applications author or a business analyst, you're on the hot seat here, since you'll need to understand which disparate sets of data need to be integrated … IT shops have a choice of technology to do the job, but the most important decision is to get started!” Computerworld also reports: “It's time to get serious and build bridges, so users can get all the information they need in one place at one time.”
If you read this sentence a few times, the question that comes to mind is why is it only now that Computerworld thinks integration is a problem? Why all the excitement when most of us have been wrestling with it for years? What has really changed here?
As we gathered around the bar, and sampled a glass of wine or two, we began to share stories about our own observations. For some time now, we had all noticed the rise of the “integration architects” and the “integration managers” – new positions within the company staffed by senior folks charged with really getting this all to work. According to some research we had done, there weren’t just 20 or 50 people claiming to have this title, but thousands! A whole new demographic within our community! And a very diverse group at that, with just as many ideas about pursuing integration as there were technologies available.
While riding the moving escalator at Heathrow airport that takes you down to the London Underground, I moved past a series of billboard posters promoting HSBC Bank. Under the banner “another point of view can sometimes open up a whole new world” there were pictures of cricket players alongside ballet dancers, with headings of Tedious? And Riveting? Another poster , immediately followed , with the same grouping of cricketers and ballerinas but this time, the headings were reversed Riveting? Tedious?, followed by the observation “a different point of view is simply the view from a place where you’re not!”
Integration, and the rise of integration architects and managers, shouldn’t be taking us by surprise any more than reading in Computerworld that it will be blazing hot topic in 2008. In the NonStop space we are beginning to see very serious deployments of Service Oriented Architecture (SOA) and for me, this has to be key in any integration planning. Integrating the data on the back-end and making sure the applications externalized as services have the right data, and it’s as fresh as it can be, is equally as important.
And where one integration architect may view SOA as the solution to modernizing legacy applications and giving them a whole new life, another integration architect may see it as just an easy way to pull new information from public web sites. Wrapping old applications with SOA, or getting access to completely new ones with SOA, doesn’t lessen the value that SOA brings to the table and while the integration architects may hold strongly differing views, they’re really not that far apart. They’re deep into solving integration issues and seriously helping the data center remain relevant in supporting today’s rapidly evolving applications.
We are never going to see the return to homogenous technology deployments centered on the offerings of just one vendor. Even where we enjoy strong relationships with a major vendor, they will be promoting different product lines and partnering with different solutions providers. We may not all enjoy the same degree of comfort undertaking application integration and not take to it as easily as we did integrating hardware, but there will be integration architects and managers within our companies aggressively pursuing integration.
As I walked through the lobby of the hotel here in London, I couldn’t help listening to the song being played. It’s by the group Maroon 5 called “Wake Up Call”, a song I particularly like. In the opening verse it says:
“And it’s not my fault, Cause you deserve What is coming now! So don’t say a word”
This has to be the anthem of every integration architect as he pursues the integration of major applications. There’s no value in maintaining silos, and if an application remains isolated, you can’t blame the integration architect. It’s not his fault. To continue with the song, its chorus asks:
“Don't you care about me anymore? Don’t you care about me? I don't think so!”
And I sure hope, that in the end, this is not the refrain of those applications left behind!