“What happened to the mobile phone charging cord?”
Every few weeks, I find myself scurrying around the house, trying to find the charging cable for my mobile phone. Actually, it’s the cable to my wife’s phone, which she usually keeps plugged in at a handy location in our kitchen. And when I find my phone’s power running low, I like to plug into that handy cable. Sometimes, my wife takes the cable to charge her phone in the car. Of course, that’s usually when I realize my phone’s battery is perilously low. Then I have to run to the other room to dig out my own cable from my backpack.
Well, you may wonder what this has to do with data centers, or you may wonder why I don’t just break down and buy another charging cable or two. The point I’m trying to make with this trivial little example is that when you share infrastructure across multiple application owners, sometimes in the heat of the immediate need you do things that unwittingly complicate life for others.
And in today’s enterprise data centers, with hundreds of applications, some of which have very complex configurations, and thousands of servers, the number of such shared connections is quite large, and the potential for trouble, especially with rapidly-changing application and infrastructure modernization initiatives, is great.
Hey, I think we’re OK, we’ve got a CMDB!
Some of you seasoned data center pros may say, hold on a minute. Isn’t this why we’ve spent years instituting change control disciplines and configuration management databases (CMDBs)? True, and these processes and databases are a great help, but sometimes the rate of change to applications and their underlying infrastructure is much faster than the rate of change in these traditional toolsets.
OK, you say, our data center housekeeping is a bit behind the curve. But that doesn’t really impact the business so it’s no big deal. Well, that may be true in the short-term. However, our experience is that it does become a big deal when you’re embarking on initiatives like data center consolidation, or modernizing applications to cloud-native platforms such as Pivotal Cloud Foundry, or modernizing your data center infrastructure.
In such situations, migrating an application from only part of the server infrastructure that it’s running on, can be risky. You can break the application, causing an interruption of a critical business process and annoying your application stakeholders. Or you can make your application configurations even more complex, with an unwieldy, hard-to-manage mix of older and newer infrastructure.
Don’t worry folks; we’re trained professionals
So you really do need to baseline the interdependencies between your applications and your infrastructure before you start on such a strategic data center initiative. But you may say, this is a big effort, and things may have changed again by the time we finish. Fortunately, today’s automated toolsets can greatly reduce the time and effort it takes to develop such a data center blueprint that documents and inventories applications as well as the underlying server infrastructure and the connections and dependencies among all of them. And keeping track of things as your migration or modernization program continues can also be done in the same automated fashion, minimizing the risks of application breakage or business interruption.
This is why Dell Services advises its clients to conduct such an automated data center blueprinting exercise as a matter of course for such initiatives as data center migration, hybrid cloud deployments, application modernization and data center infrastructure upgrades.