Guide to a cost-effective cloud migration strategy
Creating solutions to the challenges brought about by Covid-19 has led many businesses to advance their tech roadmap dramatically. An IT infrastructure now needs to securely support a remote-first workforce and enable businesses to mitigate the challenges of accessing information in uncertain times. Any previous hesitations around the adoption and costs of moving an IT infrastructure to the cloud have been put aside or worked around, as more IT decision-makers (ITDMs) see the benefit of being fully “in the cloud.”
John-David Lovelock, the vice president of research at global research and advisory firm Gartner, believes that Covid-19 has advanced investment in the cloud by several years:
"In 2020, some longer-term cloud-based transformational projects may be put on hiatus, but the overall cloud spending levels Gartner was projecting for 2023 and 2024 will now be showing up as early as 2022", Lovelock states.
A migration to the cloud allows a workforce to work remotely and businesses gain peace of mind knowing they can operate efficiently on a global scale. Developing a strategy for a successful cloud migration means greater cost efficiencies and savings over time as well as a fully tested infrastructure.
In this article on how to develop a cloud migration strategy, we will look at:
- Why are businesses moving to the cloud?
- The cloud service models: Software, Platform, Infrastructure and Everything as a Service (SaaS, PaaS IaaS and Xaas)
- What are the types of cloud migration strategy?
- How to prepare for a cloud migration?
- What are key challenges to a cloud migration strategy?
- The steps to a successful cloud migration
Before we look at what the steps to a successful cloud migration should include, let’s look at the basics.
What is cloud migration?
Cloud migration is the process of moving some or all applications from a local, on-premises data center to the public cloud.
Why are businesses moving their applications to the cloud?
Moving to the cloud offers business of all sizes benefits such as:
Hosting applications in the cloud means a business can save money as they need to only pay for hardware as the business grows. As seen in the aftermath of the seismic shift to remote working during Covid-19, being hosted in the cloud has given teams the freedom to work according to their preferred time zone, location and team size, as the cloud can be accessed anytime and anywhere - all that is needed is a connection to the internet. Depending on the type of cloud hosting model chosen, scalability and growth alongside the business needs are key advantages.
The amount users pay with each cloud hosting platform varies according to how much they consume of the services being delivered. Businesses will only pay for those that the business used, meaning cost efficiencies to boost the bottom line.
Extra computers, hardware maintenance, energy bills, dedicated IT personnel, data security - all the standard operating costs of maintaining physical servers can be significantly reduced or removed by a move to the cloud. Recovery solutions for data are in-built to many platforms meaning less additional expenses on measures to protect data from loss or equipment failure. Data encryption is inherent to hosting in the cloud with platforms automatically providing unique security settings for every user.
Agility is now central to any development process. If one server is down, teams and customers can still access data as a benefit to being hosted in the cloud, another server can be ready to take its place. Service is not impacted internally or noticed externally because additional servers automatically step in. This can mean little or no impact on any business Service Level Agreements (SLAs) in place and a reduced impact on an engineering team, who might normally need to assign resources to support a server that goes down.
Faster decision making:
In the cloud, data is accessible and available anywhere there is an internet connection, this empowers faster decision making as the most accurate and most recent data is being deployed to the infrastructure and decision makers at all times.
For an engineering team, migrating applications to the cloud benefits by allowing:
Cloud platforms support innovation and new functionality as teams only pay for the time required and can explore other cutting-edge technologies made available to them via the platform.
The customer experience is not impacted as full redundancy is in place at all times.
A continued integration pipeline means no release planning is needed and developers can work around deployments.
To support global growth for any business, local hosting in the cloud means optimal performance in new markets and the best customer experience as the data is closer to the relevant users.
Cloud computing hosts can also act on an hosting issue once detected, leading to a resolution before it causes any problems or needs escalation.
What are the types of cloud service models?
The four ways a business can use the cloud are:
Software as a Service (SaaS)This cloud service model is popular with companies that don’t want to manage server issues or short-term projects. The model uses the internet to deliver the applications and third party vendors manage the service. Businesses pay a monthly or yearly fee to this third-party.
Key advantages of SaaS as a cloud model:
Ability to run applications without maintaining and updating infrastructure.
Streamlined maintenance and support.
Access to additional cloud-based software, programs, and applications in the platform.
Main disadvantages of SaaS as a cloud model:
Being locked in to use the one vendor.
Reduced customisation abilities.
Lack of control over data and governance.
Popular SaaS hosting platforms include:
Infrastructure as a Service (IaaS)
IaaS is known as the most flexible model and allows a business to invest on demand, rather than buying a full suite of hardware from the start. It offers pay-as-you-go pricing for services such as storage, networking and virtualization via a dashboard or an application programming interface (API) meaning advanced control over the infrastructure. The business still manages applications, runtime and data.
Key advantages of IaaS as a cloud model:
Gain more computing power to run variable workloads with fewer costs.
The IaaS provider manages servers, hard drives, networking, virtualization and storage.
Increase the cloud solution's capacity anytime and purchase hardware as a business grows.
Main disadvantages of IaaS as a cloud model:
Investment in team training may be required to manage the infrastructure and be an extra cost.
The business or client must use the tools made available for data security, backup and business continuity.
The need to update and secure legacy applications prior to migration.
Popular IaaS cloud hosting platforms include:
Amazon Web Services (AWS)
Read more about our migration to AWS here.
PaaS (Platform as a Service)
PaaS is similar to SaaS, however software is delivered via a platform and the platform is delivered via the web.
Key advantages of PaaS as a cloud model:
Get faster releases of application updates.
Use a ready-made infrastructure environment to develop, test and deploy applications.
Improves the performance of a team as apps can be customized without the need for ongoing software maintenance.
The model is highly scalable as less coding is needed.
Main disadvantages of PaaS as a cloud model:
Having the business data sit with a third-party vendor carries risks
Integration with legacy applications and systems might become too burdensome to the migration process
If a new PaaS provider or models is chosen in the future, the chosen vendor need adequate migration policies in place to enable this for a business
Popular PaaS cloud hosting platforms include:
Google App Engine
Everything as a Service (XaaS)
This model is growing in popularity. It is essentially “servitisation” or combining products with services and data insights derived from Internet of things (IoT)-powered devices like mobile phones or smart devices that deliver to customers over the internet and outsource everything that is not core to a business.
What are the types of cloud migration strategy?
In 2011 Gartner outlined five different strategies for application migration to the cloud:
Lift and shift / rehost—moving applications to the cloud as-is.
Refactor—modifying applications to better support the cloud environment.
Replatform—moving applications to the cloud without major changes, but taking advantage of benefits of the cloud environment.
Rebuild—rewrite the application from scratch.
Replace—retire the application and replace it with a new cloud-native application.
How to prepare for a cloud migration?
Before undertaking a cloud migration team assess:
All existing resources and the interoperability of legacy applications prior to migration.
Decide on the right cloud solution and vendor - consider General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA) requirements that need to be met
Identify the current level of skills and if training is required.
Select an architecture for the new cloud infrastructure.
Evaluate what data needs to be migrated first, what data is non-critical and what needs encryption while in transit.
The main goal of a cloud migration is to ensure a seamless non-disruptive transition in the quickest amount of time possible - with no data-loss. Proper planning will mitigate the risk of and help prepare for all scenarios, especially any unforeseen issues that could arise.
Further consideration to include in a migration strategy is to synchronise changes in the source data during migration. Ongoing testing of performance and availability means a business can ensure their current Recovery Point Objective (RPO) and Recovery Time Objective (RTO) are being met.
What are the key challenges faced in a cloud migration?
Having unclear goals across all stakeholders.
A mismatch of people, processes, technical talent and resources before, during and after the migration.
Poor planning and incomplete understanding of existing and future application needs.
Choosing the wrong technology, cloud product and service model.
The steps to a successful cloud migration include:
Cloud Migration Checklist
- Choose a model:
Understand the differences and advantages of the four cloud service types to decide on the right one for the business.
- Choose a migration strategy:
The strategy taken will impact when the migration takes place, its duration and the ongoing work to update applications, for example, in the case that “lift and shift” is chosen or if code “refactoring” is essential to complete prior to the cloud migration.
- Prioritise applications:
Mass-migration or move elements and test as the migration progresses. Map interdependence and assign a priority to the applications to ensure the most secure accurate data is migrated at the right time and there is not only no data-loss but no loss of access for customers to their data. One tip is to migrate apps with fewer dependencies first and to test them in their new environment.
- Create a project plan:
Plan the migration according to the location of the data and the dependencies identified and mapped beforehand. Also understand what access is needed for applications and when.
- Establish security policies:
Work with the new provider to ensure their architecture satisfies the data regulation requirement in every territory operated in and for future launches. Future-proof for uncertainty, for example, data-exchange post-Brexit.
- Set KPIs and post-migration KPIs:
Categories to arrange cloud migration KPIs under include:
- User experience (Page Load Time, Session Duration)
- Application performances (Error Rates, Throughput)
- Infrastructure (CPU Usage, Disk Performance)
- Business engagement (Conversions, Conversion Rates and Engagement)
- Plan resources:
Assess what additional resources are needed or what resources needed to be relocated to deliver the integration, for example:
- Does the business need to train any or all of the engineering team?
- What additional budget is required for this?
- Will the business roadmap be impacted?
- What can be deprioritised as the migration takes place?
- Test, review, adjust:
After migrating key elements, test their success in their new environment against KPIs. Adjust and refine the plan to ensure the application architecture is successfully migrated.
With sufficient planning, the right external support along with support offered by your new cloud hosting platform provider and internal cooperation, a business can successfully migrate to the cloud and enjoy cost-savings, a competitive advantage and scalability.
CurrencyFair can help businesses save on international money transfers to pay overseas suppliers, international employees or receive payments in another currency. With competitive exchange rates, businesses of all sizes can avoid excessive transfer fees and charges and lower their FX costs with our secure multi-currency account. Experience world-class customer support - as standard.
The Best Accounting Tools for SaaS Companies
November 18, 2019
July 5, 2019
October 19, 2020