Tackle Your Cloud Challenges, One Scenario at a Time

By Pavel Despot, Sr. Product Marketing Manager, Akamai Technologies

Where are you in your cloud transformation? Whether you are beginning migration, are modernizing applications, or have been managing a cloud environment for years, chances are you are (or will end up) adopting multiple cloud platforms. According to Gartner, 75% of enterprise customers using cloud infrastructure as a service (IaaS) will adopt a deliberate multicloud strategy.

Prepare and plan to overcome challenges
With all things transformational, however, come challenges. A recent Techstrong survey identified a variety of challenges that are driving teams to multiple cloud providers; including achieving price-to-performance and infrastructure openness while overcoming complexity. Based on our experience working with cloud teams around the globe, most of the challenges we see fall into two categories: (1) migrating to cloud services, and (2) optimizing consumption of existing cloud services. It’s also not uncommon to see both of those challenges within a single organization, since teams operate with different constraints and projects. Preparation and planning are critical to overcoming these challenges.

Best practices for common cloud scenarios
A recent Forrester report may help with that effort. The Forrester Tackle Your Cloud Challenges with Forrester’s Scenario Quick Start Cards report can help start you off with some best practices for common cloud scenarios (Figure). These scenarios cover a majority of challenges teams encounter, regardless of where each one is in their cloud journey.


Figure: Cloud modernization journey

Breaking down the scenarios
Some scenarios are more common for certain levels of cloud maturity. Take cloud security, for example. Securing cloud data and workloads is generally one of the first orders of business for migrations. Per the report, however, it’s “much more complex than on-premises security and requires explicit planning and purpose-built tools.” Without that understanding, many enterprises just take the same approach that worked on-premises.
As a result, security and ops teams are left with multiple security workloads and an increase in sprawl. That duplication is a main cause of why escalating cloud costs (Scenario B), adapting operations for the cloud (Scenario E), and achieving compliance (Scenario Q) repeatedly rank as organizations’ top cloud challenges.


With multiple options for where each workload should live, you shouldn’t settle for any environment that works; you should determine where each workload executes best and most efficiently. You should also thoughtfully shop around your workloads to avoid the common pitfalls.

Defining workload shopping
Workload shopping — the process of determining the optimal environment for each workload — is an important step in architectural and infrastructure planning. It allows IT teams the opportunity to select the most appropriate platform for each component, instead of defaulting to a less efficient approach that was acceptable in previous architectures.

Workload shopping is particularly important when executing the following scenarios in Forrester’s new quick start report:
Protecting workloads in the cloud (Scenario F) — Protecting cloud compute and data resources requires a variety of security controls. Those security controls (workloads themselves) can either be deployed in each region where we have workloads running or in a unified layer at the edge. The optimal approach for security is to deploy all the controls in a single, cloud-agnostic layer. Doing so simplifies SecOps by reducing the number and types of solutions that have to be managed. It also improves the portability of the workloads themselves by decoupling security needs from compute and storage.

• Moving enterprise applications to the cloud (Scenario H) — Despite years of cloud migrations, an untold number of enterprise applications are still waiting to be migrated to a more efficient cloud environment. Based solely on the ability to execute that workload, nearly any cloud provider would be acceptable. As the report points out, though, “cloud-native services [to] enhance the application” should be considered in this new environment. That includes application security options, access control, and Zero Trust solutions. For these workloads, an environment with those integrated capabilities eliminates the need to build your own solution from various components, resulting in simpler migration and operations.

• Tackling application modernization (Scenario L) — Eventually, technical debt outweighs the cost of modernizing an application and the ROI swings in favor of action. A majority of these efforts involve monolithic architectures that need to be split into their individual components. Remember that each of these component workloads should be shopped individually to maximize the benefit of each environment. Many teams, especially those that are new to application modernization, will want to consider an environment that focuses on core infrastructure as a service and platform as a service. That simplifies the modernization and subsequent management effort of workloads subject to privacy or other regulations.

• Mitigate cloud vendor lock-in (Scenario N) — Workload portability was heralded as one of the main benefits of cloud computing. The reality, however, hasn’t lived up to the promise. There are some workloads that require features and, therefore, commitment to a single provider. The rest are generally built using virtual machines, containers, and storage. Ensuring that those workloads are portable allows you to deploy them to other clouds as risk management, cost, performance, and data sovereignty requirements dictate.

• Extending public cloud to the edge (Scenario O) — The edge’s role as a compute location has been growing as a result of the superior latency and availability it provides, but it’s not the right place for every workload. By definition, the edge is intended for distributed workloads with smaller memory and CPU footprints. Larger workloads should be deployed in more centralized clouds. Understanding the limits of the two architectures and deploying services appropriately ensures infrastructure components can scale properly and provide the expected user experience.

A great place to start
Whether you are new to cloud environments or are a seasoned multicloud veteran, you should always prioritize deliberate cloud planning and architecture to avoid many of the common issues with cloud migrations and operations. The Forrester Scenario Quick Start Report covers 18 such scenarios and includes actions for each one, making it a great place to start.

Cloudcloud migration
Comments (0)
Add Comment