When we talk about cloud migration challenges, the conversation is about a company switching its workloads from an on-premises data center to a public cloud environment. But what about cloud-to-cloud migration? The benefits of cloud-to-cloud migration Why would a company go to the trouble of migrating its entire infrastructure to the cloud, invest in one cloud service provider but turn to another? Cloud migration is nothing new. Companies have embraced cloud adoption and are becoming more accustomed to using cloud services. Now that AWS, Azure, and Google Cloud Platform currently lead the market (and other fast-growing platforms) and keep offering new and better options in terms of pricing and service, switching providers can be fruitful. Choosing a cloud provider in the first place can be a daunting task. Businesses must make choices based on a variety of factors—cost, reliability, security, and more. But even taking all factors into account, the business environment is constantly changing. Costs may become more or less important, your geographic region may evolve, which affects the cost and availability of services), and priorities may shift to the point where another platform may be more suitable. Perhaps your primary motivation for migrating to AWS a few years ago was reliability and risk mitigation. You want the gold standard when other providers keep popping up.
A few years later, productivity tools like Google's G Suite can be useful for your business. You now have business partners using other platforms like Azure or Google Cloud. You realize your needs for software have changed, business partnerships have had an impact, and it's clear that another vendor might benefit. Not to mention, cloud services themselves are constantly changing, and over time you may find better pricing, service level agreements, scalability, and higher performance with another provider. While all of this makes sense, in theory, let's look at a real-world example: GitLab case So many users have expressed dissatisfaction with Microsoft's acquisition of Github that hundreds of thousands of people have moved to another Git repository manager, GitLab. In a industry mailing list announced that they had decided to switch Microsoft Azure to another cloud provider - Google Cloud Platform. Ask GitLab's Google Cloud Platform Migration Project Lead, Andrew Newdigate, why they're migrating to GCP, and he'll probably mention that service performance, reliability, and things like Kubernetes are the future. Kubernetes is Google's first open-source project to be used for application management of multiple software containers "to enable reliability at scale." Also intriguing is that GitLab can use Google Kubernetes Engine, a service designed to simplify the operation of Kubernetes clusters, as part of its cloud migration. The use of GKE is cited as another driver for
GitLab, which wants to focus on "improving the stability of the scalability of GitLab.com by moving our workforce to Kubernetes using GKE". Sid Sijbrandij, CEO of GitLab, cited better pricing and superior performance as the reasons behind the migration. In an interview with VentureBeat, he said: “As the Google of public cloud, they have more experience than other public cloud providers because they basically build a cloud for themselves […] and you will find that their network quality is ahead of others in terms of network and so on. All. It's more reliable, less jittery, and the way they do it is really, really impressive, and we're excited to start hosting Gitlab.com." The challenges of cloud-to-cloud migration There are many factors that influence a company's decision to choose a cloud provider, and these factors don't stop once you start building your infrastructure in a particular cloud. Over time, other vendors may prove to be better suited to your business needs. But just as cloud adoption faced challenges in the first place, there are similar challenges in transitioning from cloud to cloud: data transmission. Transferring data between different cloud service providers is a complex task to say the least.
Like the data transfer from the enterprise to the cloud, the information is transferred over the internet, but between cloud providers rather than from server to cloud. This raises the question of data download speed, and as a rule of thumb, you should avoid transferring large amounts of data at once. Moving data out or into the cloud can even incur huge transfer costs. Potential downtime. Downtime is also a risk. It is important to consider inconsistencies in data, check network connectivity and prepare for the possibility of real application failure during migration. Technology to adapt to the new cloud. You built an application for Azure, but now you're using Google - it's not that simple to get it from one platform and expect it to run (with the same benefits) on another platform. Expect to spend a lot of time reconfiguring application code to take full advantage of the new platform. Control costs. This is often misunderstood or grossly underestimated when considering the time and cost of moving to the cloud. Again, the same applies to cloud-to-cloud migrations. By now, you have a better understanding of the intricacies of cloud service offerings, pricing models, and cloud adoption budgets - for the services you're using.