I recently attended Google Cloud Next ’17, and I spent a lot of time considering where Google Cloud fits into Workstate’s Cloud Shift ecosystem of helping people migrate to the cloud. While listening to the conference keynotes, I thought of an interesting way to compare AWS vs. Google vs. Azure to LEGOS, which I turned into recent blog post. I also remember feeling that Google was lagging behind in cloud thought leadership.
Much of this occurred to me as I listened to Diane Greene, SVP of Google Cloud, present her thoughts on their business to 10,000 people during the opening keynote. Diane kept using the phrase, “Lift and Shift” with respect to cloud adoption, and emphasized how invested Google is in becoming a cloud platform leader.
I noticed that the normally charged and anticipatory developer excitement around a tech leader keynote was missing: It was clear that the audience wasn’t very interested in hearing about Lift and Shift. Instead, conference attendees were there to hear about the latest-and-greatest new services from Google; and we had to wait until day two to get energized by a forward-thinking Cloud Shift keynote. It was during this 24-hour pause that I had time to consider this seeming disinterest in traditional migrations.
Did it mean Lift and Shift to the cloud was dead? You can certainly search online and find plenty of attention-grabbing headlines declaring it so. You’ll also run across headlines announcing hybrid cloud or multi-cloud as the latest in cloud. Our own Lift and Evolve cloud solution is about cloud innovation – transforming your applications to take advantage of services and techniques that weren’t available before public cloud made them possible. It’s easy to be enamored by the incredibly rapid launch of new cloud products and services, but I would attest right now: Lift and Shift – Not Dead Yet.
For most, the term Lift and Shift refers to the simplest cloud migration. The main goal is to take the current application or server workload, and move it from on-premises to a cloud-hosted version. It’s often the quickest way to get to the cloud, and to begin reaping benefits.
For all of us sitting in that keynote speech, it’s much more exciting to say, “we want to use the latest cloud techniques,” when crafting a cloud strategy, instead of focusing on the foundation of simple migration. We do this because we believe strongly in the cloud as the future of computing. But, migrating all of your apps to a cloud native architecture is difficult, and potentially not worth it. Oftentimes a simple Lift and Shift (or Refit) is the best place to start.
This probably sounds like common sense, but experience shows otherwise. I was just recently on a call listening to an enthusiastic cloud-first plan that would coincide with upgrades to the organization’s already successful application. As cloud application developers, it’s so easy to envision everything we could do to evolve and transform an application to be “cloud first.”
This is where Workstate’s experience with application development informs any development planning. The first question we asked in the aforementioned call was, “Are there any performance issues (maintenance, user experience, cost) with the current application?”
If there are no performance issues in an app, one should be very mindful before abandoning a straightforward Lift and Shift. From this “if-it-ain’t-broke-don’t-fix-it” position, you can identify cloud evolution options for strategic investment in the future of the workload.
I’ve always considered thoughtful and judicious application of technology as the best path to success. It’s not uncommon for an application to get considerable benefit from the cloud with a few, small changes during a straightforward Lift and Shift. Small changes – like improved IAM, using a fully managed cloud database, or moving state information to a high performing cloud option – can make a huge impact. Only new applications, or complete overhauls, are likely to be candidates for a completely cloud-first strategy.
Remember: Those of us building or supporting an application think most about the technology. The application users care most about how the applications performs. Your executives care about performance, too – but also, they are concerned about how much it costs to create and maintain. With that in mind, choose your cloud strategy wisely for each workload, and don’t forget that Lift and Shift is most assuredly, not dead yet.