eBook
Success in the Cloud:
5 Steps for Effective Migration
By Insight Editor / 23 May 2023 / Topics: Compliance Hybrid cloud Migration Cybersecurity
Cloud can offer scalability, automation, easier innovation and cost savings. But there are many questions to consider in a move to the cloud. Read this guide for a condensed roadmap to help increase your chance of success.
Overview
It’s safe to say the cloud is no longer “the next big thing” — most organizations are already taking advantage of the many benefits cloud has to offer. In fact, our Insight-commissioned Foundry survey of 400 decision-makers found that more than half of respondents’ data resides in public or hybrid cloud.1
Cloud may be commonplace, but success in the cloud is not automatic. Cloud migration — whether you’re just starting out or expanding your existing cloud footprint — brings with it several questions and considerations: Which applications are better in the cloud versus on-premises? How can you migrate with minimal risk and disruption? What about security and compliance in the cloud?
This guide will help answer these questions and put you on the path to maximizing the benefits of cloud.
Key cloud benefits
- Infinite scale and automation to grow applications up or out as needed
- Opportunity to innovate and “fail fast” with new applications
- Cost savings (with more measurable, monthly OpEx costs)
The lay of the land….
Before you strike out on a new project, it can be helpful to see how others have executed similar initiatives. These findings from The Path to Digital Transformation: Where Leaders Stand in 2023 will help you get a feel for where organizations like yours are succeeding with cloud — and where common challenges lie.
- 91% of organizations rely on multiple cloud providers.2
- 49% intend to increase cloud adoption via hybrid and multicloud models.3
- 20% plan to repatriate public cloud workloads to an on-premises model.4
Integration, security and performance management tie for the title of top challenge in public/hybrid cloud deployment.5
Security is the top challenge with multicloud strategies.6
The data goes to show that, though cloud is a crucial part of digital transformation, a) it’s not the right fit for every workload, and b) it can be tough to integrate, manage and secure — especially in a multicloud model.
For making the most of cloud, whether you’re looking at wholesale migration, application modernization or preparing your infrastructure for transformation efforts like adopting Artificial Intelligence (AI), the steps outlined in this ebook can help you ensure cloud success.
Note: These steps don’t have to be followed sequentially. Some may overlap, while others may be shortened, altered or streamlined to accommodate each organization’s unique schedule and priorities.
Step 1: How “cloud-ready” is your organization?
Much of cloud success comes from the initial groundwork laid during these first few phases of assessment and planning. This first step involves assessing how ready your organization is for a migration initiative. It takes a very close look at current IT operations, current processes and people. Sample questions here include:
- How do you control and manage current IT costs? How will that translate to cloud?
- What skills or training are needed to support new cloud roles?
- How does the business currently use its applications?
- What levels of security and compliance do you need in the cloud?
- What types of identity management and access controls are needed for cloud?
Step 2: How “cloud-ready” are your applications?
This second phase looks closely at how ready your applications are for a move to cloud. But it doesn’t just stop at applications. It looks carefully at the needs of application workloads. A workload represents an application and any upstream or downstream dependencies an application requires to operate successfully.
A workload may involve certain compute, network or storage needs. It can require one or more databases. It might even involve other, interconnected applications. This is where you start to determine “which workload goes where” regarding cloud. This is not always an easy task.
Insight’s survey found that more than half — 52% — of surveyed organizations with a multicloud strategy struggle with workload/platform alignment.7
At this second phase, it’s critical to ask questions like:
- What does this application depend on to operate? What base metrics can I collect to learn more? How much does this application workload cost to run now?
- If this workload moved to a public cloud, how often would users need to send data to it (or, more importantly, receive data from it)? How much data? How does that translate into cloud ingress/egress charges?
- What “cloud-to-on-premises” connectivity options and performance levels are needed for this workload in the cloud?
Answers to these questions help you prioritize which application workloads are more cloud-ready than others. It can also help determine any cost investments toward application modernization or optimization that may be required before certain legacy workloads can succeed in the cloud. You may find some workloads better suited for hybrid cloud or continued, on-premises operations — at least for the short term.
Moving workloads: Who, what, where, when…
It’s important to ask the right questions when deciding which workloads should go where. A good start is the reporter’s fact-finding mantra:
- Who? Who currently uses this application workload?
- What? What upstream/downstream dependencies does the workload have with other hardware or software?
- Where? Where do the workload’s components currently reside (physical servers, VMs, etc.)? Where (as in the specific cloud provider and service) do you plan to move it?
- When? When did you hope this workload could be migrated to the cloud?
- Why? Why is this workload a good candidate for the cloud? Conversely, why is it NOT a good candidate? Why does a specific cloud provider offer the best fit for this workload?
- How? How, more specifically, do you plan to move this workload?
Pro tip:
When moving to the cloud, don’t just do the bare minimum. Hasty “lift-and-shift” moves to the cloud lead to unexpected charges if you estimate your resource requirements incorrectly. It can also be quite costly to retrofit later if problems arise. Take time to plan necessary workload resources accordingly before moving them to the cloud. Similarly, plan an extra buffer of cloud resources per workload for unexpected peaks in use.
Step 3: Design and build your cloud.
This phase takes all of the data and findings you’ve amassed and develops a detailed action plan to move forward with the cloud. Take the time to work out the details here. Done properly, these plans will serve as a critical roadmap to guide you smoothly toward success with cloud services. At this stage, you will actively:
- Architect a cloud-based design that aligns with your business objectives.
- Identify and start to incorporate any needs for cloud backup or disaster recovery.
- Outline a viable hybrid cloud migration strategy, where needed.
- Evaluate and start to select key cloud providers and services.
- Start setting up the foundation of cloud services. This may involve starting to build sites or setting up and exploring subscription accounts and how they can be managed.
- Start implementing appropriate storage, compute and connectivity options.
- Implement any necessary services based on Platform as a Service (PaaS) or Software as a Service (SaaS) cloud service models.
- Start to document and communicate cloud plans, in detail, for both current and future workloads.
Step 4: Make a trial run.
By this point, you should have done much of the initial groundwork. Now it’s time to try out your plan and kick some tires. Two core activities are typically involved in this stage: Proof of Concept (PoC) and migration of non-production workloads.
The PoC stage
This is the perfect time to try one or more PoCs. Want to see how one Cloud Management Platform (CMP) compares to another? What about how a few sample virtual machine workloads work in the public cloud? You might want to explore how a cloud-based Database as a Service (DBaaS) platform like Microsoft® Azure® SQL could support workloads in contrast to your on-premises SQL-based environment. This is the perfect time to try things out. At the PoC stage, you can start to:
- Build and test processes (prior to building all regions and services).
- Deploy representative workloads.
- Assess for any added costs or capabilities.
- Validate functionalities.
- Answer other questions, such as how to integrate logging systems or ways to manage consumption of cloud-based services.
Migrating non-production workloads
This is where you start to actively migrate a good portion of your identified, nonproduction application workloads to the cloud. This is the chance for a better trial run to see how production workloads might run in the cloud.
For bigger organizations with several hundred (or even several thousand) workloads, this process often occurs in phases. Phase 1, for instance, might involve migrating only certain identified workloads, move groups and processes. This includes migrating or replicating any data needed to support the workloads. Included is also deployment of important backup and disaster recovery processes to support the migration.
More exploratory questions often crop up here as well, such as:
- How will all those virtual machines be logged in my Change Management Database System (CMDB)?
- How will my support team manage these workloads?
- If something crashes, how will logs be sent to a centralized logging system?
- How should we update support-related documentation, service catalogs, runbooks or support contact information to reflect services now running in the cloud?
- How should we rethink Service-Level Agreements (SLAs) when workloads are in the cloud?
- How are the workloads performing? How are cloud resources being used?
- What adjustments can we make to better control, manage, or optimize workloads for security, governance and cost?
Pro tip: Cloud migration for the risk-averse and the risk-takers.
Many clients prefer to go gradually to the cloud by first migrating manageable clusters of low-risk, non-production workloads, and then making adjustments as they go. A few others are willing to take more risk — they may choose to migrate as much as thousands of workloads on a more aggressive timeline. The choice depends on how much risk you are willing to incur.
Step 5: Going live
All steps leading to this point have prepared you to now go live and migrate targeted production data and workloads to the cloud. By this stage, you should have a good handle on new roles and who has them, and how workloads will be secured, managed and supported. Testing and validation are still required here, just as when migrating non-production workloads.
In case there are challenges or sudden changes in move schedules, established migration runbooks with rollback procedures should also be available for all parties involved in the migration.
Here as well, you will:
- Establish move groups and processes.
- Test and validate.
- Migrate or replicate workload data.
- Migrate applications based upon earlier successful migrations with non-production workloads.
By doing the earlier planning and migration groundwork with non-production workloads, you should have higher confidence that this step of going live will go smoothly.
Conclusion
The cloud has become an essential building block in many organizations’ growth strategies. Not only does it deliver on the benefits discussed in the beginning of this guide, but it often serves as the foundation for more advanced transformation, including application modernization, AI adoption and more.
Navigating cloud for success is an ongoing journey requiring strategic preparation, cautious implementation and consistent evaluation at every step.
Explore the resources below to learn more about what it looks like to start your cloud journey with Insight and see the kind of results we can help you achieve.
Sources:
1 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 11. Commissioned by Insight.
2 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 11. Commissioned by Insight.
3 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 21. Commissioned by Insight.
4 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 11. Commissioned by Insight.
5 Ibid.
6 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 43. Commissioned by Insight.
7 MarketPulse Research by Foundry Research Services. (February 2023). The Path to Digital Transformation: Where Leaders Stand in 2023. Slide 43. Commissioned by Insight.
