The COVID-19 Rush to the Cloud Results in Steep Costs and Chaos

For many companies, their data center capacity was not built for the instant tsunami-sized jolt of increased load caused by the global pandemic. Not surprisingly, the majority of companies had little choice when the pandemic hit. Procurement of new on-premises hardware in the timeframe necessary wasn’t practical, nor in their long-term plans. The desire was always to move more workloads to the cloud, but not overnight. They were forced to rush applications to the cloud.

And that was only the beginning. The pandemic continues to have an epic impact on every aspect of how we live, work, and play. These societal changes continue to accelerate the demand for cloud-based technologies. For companies, governments, and more, the need is urgent to embrace digital transformation.

But a knee-jerk change in the world does not come without digital cost and risk. Case in point, we found that as the Global 2000 accelerated the movement of legacy applications to the cloud due to COVID-19, two thirds (66%) are ‘at risk’ for unexpected cost/performance issues as a result of having simply “lifted and shifted” applications instead of optimizing prior to migrating them. The remaining one third avoided these issues by right-sizing their workloads, modernizing applications by upgrading them or rearchitecting them to run in the cloud, and decommissioning unnecessary workloads prior to the migration. This analysis is based on hundreds of Global 2000 actual multi-cloud performance and cost metrics measured.

Organizations that did not optimize workloads and infrastructure prior to migrating to the cloud used what we deem as ‘blind provisioning.’ It is a ‘fingers crossed’ mentality with an assumption that legacy applications will work fine and that workloads will move seamlessly. Some of the inadvertent errors of blind provisioning teach us to assume nothing because cloud errors can have exponential costs. This is confirmed by Gartner Research, they substantiate that organizations with little or no cloud cost optimization plans overspent on cloud services by up to 70%.

If executed with clear vision into workloads running in the private cloud, migration to the public cloud can empower a business to enter new markets, automate processes, and tap into the power of data to create a competitive edge. It is critical you know what is running on-premises prior to the public cloud shift.

Otherwise, you are subject to the dangers of a blind provisioning ‘lift and shift’ method that can result in the following issues: 1) unexpectedly high cloud provider bill with egress fees for moving workloads/applications in and out of the public cloud, 2) many legacy applications simply won’t work as expected in the cloud and 3) business processes can suffer or experience downtime as a result.

Cloud-Smart versus Cloud-First

In analyzing companies’ cloud migration strategies, we found that organizations used either a “cloud-first” or “cloud-smart” strategy. A cloud-first strategy grants broad authority to adopt cloud-based solutions, but only works well for workloads and applications specifically designed for the cloud. A cloud-smart strategy takes a more thoughtful approach, leveraging practical guidance to actualize the full potential of cloud-based technologies while optimizing for cost, performance, and risk. Most companies executing a cloud-smart strategy leverage partners to assist with the journey.

An example of a thoughtful and effective strategy is PayPal. It has effectively automated many of their cloud services, kept IT costs down, added no additional dedicated staff to manage cloud deployments – even as they scale and grow.

Those applying ‘cloud-smart,’ techniques and tools gained visibility into the following:

• Application discovery and dependencies
• Workload profile
• Correct instance sizing
• Workload replay in the cloud to validate performance and cost before migrating

An early adopter of a ‘cloud-smart’ strategy was the Federal Reserve. In its June 24, 2019 Federal Cloud Computing Strategy report, it noted: “Cloud Smart equips agencies with actionable information and recommendations gleaned from some of the country’s most impactful public and private sector use cases.”

A cloud-smart strategy allows for a deep understanding of what is actually happening in your environment. It includes real-time monitoring to detect and alert you to large shifts in performance or capacity as they happen. Cloud-smart management is not a one-time event but rather an ongoing and necessary strategy to continuously optimize resources and manage budgets. Further advantages of a ‘cloud-smart’ approach include full-stack infrastructure visibility for 1) app-centric visibility, 2) meaningful metrics for data in real time, 3) the ability to immediately eliminate bottlenecks to avoid downtime and outages, balance workloads, and 4) manage capacity to scale and grow.

How to get ‘cloud-smart’:

• Workloads optimization prior to cloud migration – discovery and dependencies: Traditional infrastructure monitoring tools don’t understand your workloads. Plus, traditional application performance monitoring (APM) tools can’t diagnose infrastructure issues. It is like two people speaking different languages.
• This creates a problematic monitoring gap. APMs focus on behavior of the application, its code, and supporting runtime environments. But, because they are largely blind beyond that—for example, they have no visibility into the storage network or storage arrays, they can only identify the simplest infrastructure problems. There is a need to see real-time performance and availability metrics correlated across the entire infrastructure from application to storage.  In doing so, you can quickly identify a problem and uncover the culprit.
• Infrastructure optimization prior to cloud migration – workload profiling, correct instance sizing: On-premises infrastructure silos are a perennial problem. Monitoring of storage, networks, systems and workloads all happening in their own worlds without reference to each other is a constant challenge. This often leads to slow mean time to repair for resolution.  Or problems with endless war-rooms and finger pointing between groups, as no single organization has a full view of the infrastructure to track down the problem. The “single-pane-of-glass” using full-stack visibility, combined with AIOps analytics to automatically track-down root cause, can identify the solution to the problem and can make these kinds of scenarios extinct.
• Risk management in cloud optimization – workload replay in cloud to validate performance before migrating: While cost is a significant consideration, it shouldn’t be the only one. If your company’s risk tolerance is very low for a particular mission-critical workload, you may be willing to spend a bit more to protect performance. The second factor is future expectations. Of course, it’s impossible to perfectly predict future requirements—especially in the current climate of transformational disruption. But understanding how much headroom each option provides, and then conducting a ‘what-if analysis’ empowers you to make informed decisions when changes happen.

Our analysis included the fact that the world was moving to a hybrid cloud strategy long before the pandemic hit:

• 80% of large enterprises initiated their hybrid cloud migrations at least three years ago
• 10% of planned workloads had been migrated due to caution before COVID-19

But, COVID has sped up migration. Pre-COVID, five years was the projected time for 50-60% of workloads to be migrated. Now that has shrunk to two years.

Gartner research has also substantiated our metrics of Global 2000 cloud migration growth. The firm reported that by 2025:

• 85% of large organizations will have engaged external service providers to migrate workloads to the cloud, up from 43% in 2019
• 85% of large organizations will use external service providers to migrate data/storage to the cloud
• 80% of organizations will use external service providers to move an application/compute infrastructure from one environment to a different cloud environment

The public cloud migration rushes during the past few months have exposed the need for cloud cost optimization broadly. For their most important workloads, organizations are now well aware that they need an investment in re-platforming the application to take advantage of cloud efficiencies. Companies that created and executed a cloud migration optimization plan were able to avoid increased cost and degradation in performance of workloads. In essence, they were cloud-smart.

We all understand that COVID will continue to upend business. For IT, and cloud, it can be a very positive driver of change. The takeaway is that every sized enterprise needs to slow down and take the time to assess current legacy systems, project how to improve cost and performance for any style cloud deployment, and simulate potential issues and their solutions BEFORE migrating to the cloud.

Infrastructure always needs to be refreshed and it can be done confidently with the right approach. Just, make sure you are not flying blind.

By Scott Leatherman