Silver linings through cloud rightsizing

Amid the past decade’s cloud frenzy, some companies’ cloud migration strategies are returning to earth. While migrating certain systems and workloads to the cloud offers scalability, efficiency and on-demand resource availability, cloud migration is not one-size-fits-all. During a massive two-year cloud migration project for a global internet carrier, I learned the importance of cloud “rightsizing” when pursuing ambitious cloud migration. Many companies can benefit from cloud transformation, but it’s essential to integrate a cloud rightsizing strategy to appropriately scale your migration according to your organization’s IT system requirements.

Cloud rightsizing: how much is enough?

So, what exactly is cloud rightsizing? It’s a strategy that describes optimizing cloud resources (such as compute, storage and networking) according to the needs of applications and workloads running in an organization’s cloud environments. The goal is to enhance cost efficiency and resource utilization, helping companies maximize their cloud migration benefits while reducing management complexities and costs.

Other considerations include best practices concerning which workloads and systems should be stored in the cloud and which should be left in on-premises servers. While many companies initially dream of migrating nearly 100 percent of their systems to the cloud, it’s often unrealistic due to technical, legal and operational constraints. While it was a vast undertaking, our own cloud migration taught us several cloud rightsizing considerations for any organization pursuing a similar transformation.

The telco cloud journey: top migration considerations

When we began our cloud journey, we were undergoing a massive separation from Telia Company. This included moving to new offices, building a new network operations center (NOC), migrating and splitting contracts, completing a rebranding, creating new finance functions and much more. In addition to time-consuming processes like changing the names of legal entities across the globe (and ordering lots of new office furniture), this separation entailed migrating approximately 250 IT systems out of Telia and into Arelion’s own IT environments. Suddenly, we needed to choose the most flexible option to complete this challenging system migration, and we had to do it in 24 months.

We faced a fork in the road. We could build many on-premises servers to house these IT systems, or store more of our systems and applications in the cloud. We set a goal ratio of cloud versus on-premises hosting, hoping to migrate most of our systems into the cloud. However, that was just the beginning of our journey. We reached a 60/40 split after our cloud migration, with 60 percent of our systems in the cloud. This will be continuously challenged and carefully evaluated when new systems arrive, but we no longer see storing nearly 100 percent of our systems in the cloud as a realistic goal.

As a positive side effect of this project, we conducted a quick evaluation and discovered that we no longer needed approximately 30 systems which we previously had for legacy reasons, so we were able to discard those and make our operations leaner. We also had systems, like Salesforce, which were already in the cloud and did not require migration from Telia’s environments. These are crucial aspects for companies to consider. Can they mitigate IT system bloat by eliminating certain legacy systems? Should they have their own system in the cloud, or just buy a system that is already available online? Over the course of this project, we learned that data sensitivity, expertise, scalability and cost are the top cloud migration considerations.

Insights from behind the curtain

Data Sensitivity and Security: We ambitiously planned to migrate most of our IT infrastructure into cloud-hosted environments. However, this proved unrealistic as the project progressed. While exploring our IT system requirements, we discovered our vital legacy systems could not reside in cloud environments due to their operational constraints. Security and data sensitivity were primary constraints for certain systems driving critical network resources and financial functions, so we kept these in on-premises environments. To a lesser extent, some systems had such old code that they weren’t a good fit for the cloud, meaning a simple server running on an old operating system was still the best, most cost-effective solution for legacy systems vital to our business operations. These are essential questions for companies to ask themselves

• How much of your data can you store in the cloud, and how much is too sensitive to migrate?
• Where is your cloud physically located, and how do local laws and data protection regulations impact your data sovereignty?

Third-Party Expertise: We had many talented engineers who ran our systems efficiently, but we didn’t have anyone with experience moving those systems to cloud environments. So, we hired a consultant and a permanent employee with cloud migration expertise. This strategy saved us immense costs and time because these specialists knew the best questions to ask potential cloud providers, and they understood the benefits and drawbacks of various cloud options when mapped to our IT system requirements. These questions can cover a broad range of technical infrastructure, security, billing and migration-specific topics, including:

• What are your disaster recovery capabilities and data backup procedures?
• Are there hidden costs for data transfer, storage or customer support?
• How do you handle legacy systems during migration?

Scalability: Aside from cost reduction, scalability through dynamic resource allocation is the primary benefit for any company undertaking cloud migration, whether they’re an enterprise organization, data center operator or global network operator. In our case, we needed to maximize flexibility to ensure we could adjust resources according to real-time demand and fluctuating workloads in our internal IT systems. This helped us enhance our business agility in responding to operational demands while ensuring our internal resources were not overprovisioned or underutilized.

Striking this balance in resource allocation also reduces the risk of internal downtime or outages that impact customer experiences in an always-on business world. Minimized downtime risk and reliable scalability across global regions is vital to the nature of our customer service strategy, so our ultimate choice reflected this business need. When exploring the scalability and flexibility benefits offered by a cloud provider, companies can ask these questions:

• What are your procedures for ensuring dynamic resource allocation through upscaling and downscaling?
• How do you scale resources during peak times while minimizing operational downtime?
• How do you support load balancing and failover processes?
• What is your approach to geographic scalability and data residency?

Actual Cost Savings Versus Expected Cost Savings: Cost reduction is the driving force behind many cloud transformations. However, contrary to most popular rhetoric, we haven’t realized significant cost savings from our cloud migration just yet. But we have enjoyed flexibility, predictability and resource availability benefits that justify our migration. We can always predict the monthly cloud-hosting costs, and our cloud provider is always available to quickly add on-demand storage or computing power when needed. These are the true silver linings of our immense project.

The difficult truth is that you may not immediately see short-term cost savings just because your systems are in the cloud. But that doesn’t mean you shouldn’t integrate cloud services into certain parts of your IT infrastructure strategy, either. To maximize their digital transformation, companies must learn how cloud-hosted IT systems address their business needs and then weigh those benefits with their migration’s upfront and ongoing costs. Short-term cost reduction is a potential benefit, but it should not be assumed.

Striking a delicate balance through cloud rightsizing

Companies pursuing digital transformation should temper their expectations with the reality of their IT system requirements. By applying these lessons to their cloud rightsizing strategy, companies can optimize their cloud spend and secure the most value from their large-scale migration. As with many technological transformations, one size does not fit all, so it’s vital to strike a delicate balance informed by comprehensive research. These lessons gained through cloud rightsizing apply to any organization seeking to streamline digital operations, ensuring companies stay grounded instead of chasing pie-in-the-sky goals while pursuing massive cloud migration.

Mattias Fridström, Chief Evangelist