From the course: Requirement-Driven Cloud Selection and Deployment

Legacy applications

From the course: Requirement-Driven Cloud Selection and Deployment

Start my 1-month free trial

Legacy applications

“

- [Instructor] Keep in mind that as we understand our as is requirements that we're going to need to understand legacy platforms. That's where our existing as is state is, on-premise in a traditional data center. So we have lots of different platforms out there, but typically most enterprises may have Mainframe computers, which have been around for 20, 30 years, minicomputers around for 20 years, and even client-server-based systems and distributed systems are those platforms such as Linus, Apache, MySQL and Python are what we call LAMP stack. So when we do an as-is assessment, we're trying to figure out who, what, where, and when. Who is the owner? What's the security? What sorts of policies they have supporting governance? You need to understand the type of application, the databases that are attached to those applications, the different programming languages that are leveraged, and ultimately the operating systems, CPUs, in essence the platforms that they're leveraging. Also where? We need to understand what is the target cloud platform that's in front of us. If it's a database, are we moving to a cloud-native database in a public cloud provider? If it's a storage system, are we moving to a storage analog in a public cloud? Then we have to figure out the when or what happens. That means planning the migration. What's going to happen when? And which applications are going to be migrated and how? And also important, we have to figure out how it's going to get paid for? Migrating to the cloud is very expensive, and we have to figure out where the funding's coming from, making sure that we have the budget to complete the project. So keep in mind that this is fairly complex. We have native cloud services, storage, compute, databases, security, everything that kind of makes up everything in a public cloud. Then we have to figure out how to make the systems connect to data that's within our legacy system, services, the ability to talk to user interfaces, and certain gateways. So the end state that we're looking for is really going to be a mix of legacy systems, many of which just won't go away, and new cloud-based systems, many of which still need to play well and communicate with legacy systems that may remain on-premise. So this is not necessarily you moving everything within the data center to the cloud. This is you understanding the requirements to come up with an architecture that's optimized, that's typically going to balance processing between public cloud-based systems and legacy systems.

Contents