Learn more about 2 of the 7 Rs of application migration – rehosting (lift and shift) and replatforming (link-tinker-shift).
In my last blog post, I outlined the 7 Rs of application migration and the importance of infrastructure to your success. In this post, I’d like to focus on two of those seven options: rehosting (lift and shift) and replatforming (lift-tinker-shift).
I understand that how you migrate to the cloud is a huge decision and that examining each application and weighing each option is time consuming and maybe a little confusing. So I’d like to offer my perspective on meeting this challenge head on and making the best choices for your business. These choices depend on several factors, one of which is how quickly you need to make this move.
Time might be the single biggest factor in your decision. For example, if your data center is closing and your new premises don’t have space to house infrastructure, you must move quickly to store your data off premises.
Or perhaps your business is experiencing a period of very high growth, and your applications are reaching the limits of your current infrastructure. Again, speed is essential to ensure that growth continues.
Or maybe your hardware is reaching its end of life, and this is the time to decide between the cloud and investing in all-new on-premises infrastructure.
These are just a few of the scenarios you may be facing when speed of migration tops your list. Your deadline looms, and you need to make a decision right now. How best to invest in your future operations?
The fastest and simplest way to move an enterprise workload to the cloud on a tight deadline is to rehost it, also known as “lift and shift.” This option requires the fewest resources and minimizes the disruption of existing application workflows.
On the surface, this can look like the best option when you’re working to a tight deadline. But let’s look a little deeper, because lift and shift is a bit of a misnomer. In reality, there is almost no situation in which you would simply lift and shift without making any changes at all. Something is going to have to change.
So the real options are:
The question is not whether to tinker, but when. Lift and shift, make sure it works, then optimize? Or improve things from day one as you start to lift and shift?
This is a balancing game of risk and reward. To obtain the fastest TTM, tinkering is minimized. To optimize the migration, you extend TTM to allow more tinkering—and the corresponding innovation, scale, and cost benefits.
The more flexibility you have with TTM, the more you should be willing to tinker with the application and replatform to optimize your outcome.
What drives tinkering? Here are some scenarios:
With a cloud migration, performance will change. However, if you simply lift and shift, it’s possible that the change could be negative. It’s important to examine your performance needs and to understand the expected performance in the cloud. Performance consists of throughput and latency; the higher the latency, the lower the throughput. Performance impact may be a reason to increase the amount of tinkering you decide to do.
When an application is running in the cloud, some workloads can experience latency issues that did not exist on premises. High levels of latency negatively impact the customer’s experience and can create significant bottlenecks for important workloads. Latency in the cloud is impacted by several factors in the network topology, including distance, type of connection, the number of hops within your network, your storage type, the data location, and the performance of the file OS. Further optimization for specific workloads requires local caching of the data in specific locations.
One of the benefits of moving to the cloud is the ability to scale your operation without continuing to invest in additional on-premises infrastructure. But the benefits extend well beyond replacing a small machine with a bigger one or replacing one machine with two. The cloud enables you to scale rapidly to meet demand without the constraints of on-premises systems. But you’ll want to be able to take advantage of autoscaling capabilities (up and down) and flexible data management services, and that means more than a simple lift and shift.
Reducing IT infrastructure cost is often a major reason for businesses to migrate to the cloud. However, migrating to the cloud can be a double-edged sword when it comes to cost. For example, cloud scalability can quickly escalate costs if additional resources are not in place. Cost optimization requires agility so that you can quickly react to changes in demand without impacting performance. And agility requires tinkering.
It’s important to take backup and compliance into consideration when choosing to migrate to the cloud. One way of easing into the cloud is to shift backup loads first, while leaving your day-to-day operations on premises. But once you migrate to the cloud, backups are just as important as they were on premises. The good news is that cloud backups are easy, even when you have elected to simply lift and shift.
With just a little bit of tinkering, you can verify that your cloud application is meeting the privacy and data retention policies set by global regulations, including GDPR and CCPA. Since you don’t control the infrastructure when you are operating in the cloud, compliance becomes an important issue to consider.
Attempting a simple lift and shift may bring you face to face with the looming skillset shortage, as your organization suddenly needs people with new skillsets to manage your applications in the cloud. Just because your team can manage applications on premises doesn’t mean that they have the cloud ops skills to do that job in the cloud. Updates and management in the cloud could present significant challenges.
If you tinker with your application from the outset, you might be able to reduce your dependency on resources and leverage services provided by specialists instead of finding a way to do it yourself. Choosing a fully managed service may be your key to success. A common example is choosing to use a database as a service rather than operating your own database. On the storage side, choosing a fully managed option is a good example.
Whether you are rehosting or replatforming, NetApp has a solution for you. Let’s take a look at some of our industry-leading solutions.
NetApp® Cloud Volumes ONTAP® empowers you to lift and shift, combining the industry-leading storage operating system, NetApp ONTAP, with the benefits of the cloud. This is an excellent choice if you have relied on ONTAP on premises and want to expand its extensive storage and data management capabilities to the cloud. You can be confident that all your automations and security choices will work. Cloud Volumes ONTAP works seamlessly with your on-premises choices, and NetApp Snapshot™ technology provides fast, effective, and incremental transfer of data to cloud storage. You can ensure migration success by running your application in the cloud in parallel with on premises. Cloud Volumes ONTAP provides an amazingly fast TTM.
Replatforming is similar to rehosting in that it does not change the core architecture of your application. Replatforming involves a moderate amount of tinkering, with no need for new architecture. This option enables you to benefit from the implementation speed of the rehost option, combined with some cloud optimizations.
For example, the Azure NetApp Files service in Azure and our Cloud Volumes Service in GCS and AWS provide the unparalleled high performance and ultra-low latency you need to run even the most demanding applications in the cloud—applications like SAP, HPC, VDI, and Oracle and other complex databases.
In addition, these cloud-native services provide scalability through volume sizing and performance tiers, enabling you to scale up and down quickly and easily to meet the demands of your workloads. There’s no need to overprovision VMs to meet your peak performance requirements when you can scale up and down on demand.
The NetApp Cloud Backup service powered by NetApp Cloud Manager offers simple, efficient, and secure backup and restore capabilities to protect and archive both cloud and on-premises data to the cloud. This fully managed backup and disaster recovery service offers the reliability and scalability benefits you need while also allowing you to reduce your costs. Cloud Backup stores your data in inexpensive yet highly durable cloud-based object storage. Your data is backed up incrementally, meaning that only new blocks are backed up, and automated Snapshot data images allow you to restore to any required point in time. Your data is encrypted both at rest and in flight for end-to-end protection.
NetApp Global File Cache enables you to consolidate silos of distributed file servers into one cohesive global storage footprint, creating a globally accessible file system in your choice of private or public cloud. Global File Cache overcomes latency issues, eliminating wait times for your large files by caching active datasets to distributed offices globally. With this software-based solution, files always act as if they were stored locally. With today’s emphasis on remote working, Global File Cache has become increasingly popular, boosting productivity for your distributed workforce while reducing IT management time—and reducing costs by up to 70%.
When it comes to cost optimization, you’ll want to take a look at the Spot by NetApp portfolio of cloud services. No matter how you migrate to the cloud, you’ll need to tinker with compute and storage, but Spot can significantly help in reducing your costs. Instead of writing to a predefined compute, if you write to the Spot APIs, you gain both continuous optimization and higher performance, without breaking the bank. Spot optimizes both compute-to-cost ratio and performance-to-cost ratio.
Spot uses artificial intelligence, machine learning, and automation to automatically provide the most available and scalable cloud infrastructure for your applications. With Spot you can reduce complexity and eliminate overprovisioning and skyrocketing cloud costs.
It’s important to analyze your organization’s business needs to determine whether to rehost or replatform. If your application can be moved to the cloud “as is” without compromising efficiency or functionality after migration, you may decide to rehost. If your application requires enhancements to optimize costs or to improve resource allocation, or if it simply cannot be rehosted, you will probably choose to replatform.
In my next blog post, I will explore the rest of the list of the 7 Rs of application migration, taking a closer look at why you might choose to refactor or to write modern cloud applications.
Whichever route you choose, NetApp can meet your needs, combining our trusted data management and storage technology with cloud expertise in any cloud—or clouds—you choose.
Let us help you discover the best migration path for your business.
We’re here to help.
Get started with NetApp Cloud Services today.
Ronen Schwartz is senior vice president and general manager of NetApp Cloud Volumes business. He is responsible for accelerating the Cloud Volumes business, building the robust Cloud Volumes platform across hybrid clouds and multiclouds, and expanding NetApp cloud data management services.
Before joining NetApp, Schwartz was executive vice president in charge of technology and ecosystem, operations, and strategy at Informatica, and senior vice president and general manager of Informatica Cloud. In those roles he led product strategy, product management, and go-to-market strategy. During his 14 years at Informatica, he helped lead the company through its own transformation from on premises to the public cloud, from license revenue to recurring revenue.