Disaster Recovery

Want to fail your cloud migration? Then don't read this

Discover how to migrate your applications to the cloud using PaaS services: strategies, advantages, cost reduction and greater agility for your company.


Migrating applications to the cloud is a priority in any modern enterprise's technology strategy, but the real challenge is how to do it efficiently and securely. There are multiple paths to moving your SAP systems from an on-premise environment to the cloud, and choosing the best option can make the difference between success and failure.

In this article, we explain the 7 main strategies for migrating applications, with a special focus on the use of platform-managed services (PaaS), their advantages, and how they can help you optimize resources, reduce costs, and improve the agility of your business.

IMG RRSS (ING) - BLOG - Unikal - BMS

The 7 R's

Taking Amazon Web Services as a reference, 7 different options for moving on-premises ecosystems are outlined.

1. Retire

Decommission the application as it does not respond to any current or future business need.

2. Retain

Keep the application on On Premise, either because it is impossible to upload it to the cloud or because it is not worth the investment to migrate it.

3. Repursache

Replace the application with a different one, so there is no application to migrate.

Being purists with the concept of migration, the first three options described above do not imply a migration in themselves. The remaining 4 R's are:

4. Rehost

Also known as Lift & Shift, it consists of performing a "twin" deployment of the on-premises ecosystem we have using the infrastructure services (IaaS) offered by the public cloud provider.

5. Relocate

This is similar to the previous one, but with an important difference: if we already use an on-premise virtualization solution such as VMWare, we can migrate directly to VMWare in the cloud, minimizing changes and smoothing the learning curve.

6. Replatform

This involves making changes to the application stack when migrating to the cloud, replacing layers with managed services (PaaS) to gain efficiency and reduce maintenance.

7. Refactor

Involves redesigning the application architecture to make it more agile, scalable, and adaptable, such as moving from a monolithic model to microservices in a serverless architecture.

Replatform VS Rehost as the first phase in the migration strategy to the public cloud.

When a company considers migrating one or more applications to the cloud, it normally carries out an inventory and analysis phase of application criticality, dependencies, etc., followed by an evaluation phase of the process to be followed so that there is no impact on the business, which ends with a migration strategy.

In many cases, in order to minimize risks, the first phase of this strategy is to opt for the first of the 7 R's (Rehost) by taking the application as it is on-premises to the cloud (also known as lift & shift).

However, although it is a valid option, in many cases it is not an optimal option, especially if we take into account certain factors such as the following:

Resource sizing and elasticity 2.

What does resource sizing imply in an on-premise environment?

In physical servers, even virtualized ones, capacity is limited to the available hardware (vCPU, RAM, and disk). Since the workload is not always constant, the infrastructure is usually sized with peak demand in mind, which ensures stability, but generates an oversized ecosystem most of the time.

How does this model affect non-productive or sporadic use environments?

In these environments, even though they may be shut down on nights, weekends, or vacations, it is necessary to keep computing resources available. This generates a cost inefficiency if you migrate to the cloud with virtual machines sized at peak.

What problems does this approach bring when migrating with Rehost?

By replicating the traditional model in the cloud under IaaS, you pay for peak capacity that is not used all the time. This leads to an unnecessary expense that is poorly aligned with the elasticity that the cloud actually offers.

How does Replatform solve oversizing?

With Replatform it is possible:

  • Schedule resizing to optimize costs.
  • Implement auto-scaling groups that adjust computing resources according to real demand.
  • Bet on serverless services, where the provider allocates or releases resources automatically according to application usage, thus optimizing the ecosystem without manual intervention.

2. Licensing

Why is licensing a critical factor in the migration to the cloud?

The licensing of the technology stack (virtualization, operating system, database, and middleware) accounts for a significant portion of IT costs. In many cases, this licensing is directly linked to compute resources (such as databases) and is therefore also sized at peak.

What are the consequences of this approach when applying Rehost?

  • Licenses must be purchased to cover the peak usage, even though the resources consumed are usually much lower.
  • This implies excessive license acquisition and maintenance costs.
  • In non-productive or limited-use environments, licenses are paid for even when applications are not active.

How does Replatform optimize licensing costs?

  • It allows the use of managed services on a pay-per-use basis, with licenses included in the hourly price.
  • In this model, the cost covers licensing, support, and infrastructure.
  • You only pay for the time the environment is on, generating significant savings in environments that do not require 24/7 availability.

What risks are avoided by opting for managed services?

With Replatform and managed services, the risk of non-compliance with licensing policies is eliminated, since:

  • You always pay for what is actually deployed.
  • You avoid audit problems due to excess licenses or activation of non-contracted options.

3. High Availability and Disaster Recovery

What happens with high availability in a Rehost migration?

In Rehost, the infrastructure is moved as-is, which means that ensuring high availability and disaster recovery (DR) depends on the customer. This requires additional investment in hardware, licenses, and management, which can make the project more expensive.

What are the advantages of Replatform's managed services?

Managed services from public cloud providers typically include:

  • High availability by default, with no additional costs.
  • Disaster recovery (DR) enables execution in another region in case of failure.
  • Services aimed at reducing management, maintenance, and operation, so the customer focuses solely on using the service.

How does this translate into business benefits?

  • Less downtime thanks to the high SLAs offered by public cloud providers.
  • Less operational effort for internal teams, who can focus on innovation and business value.
  • Greater resilience and security against disruptions.

Conclusion

We have seen that the real issue is to perform the migration in the optimal way for the business, and we usually gain that optimization by opting for the option of re-platforming our application, whenever possible, for the different reasons mentioned above.

At Unikal, we believe that the ideal is to analyze the current ecosystem to validate the best way to make the migration a success, and the managed platform services (PaaS) enable the business to do more with less, optimizing both the direct and indirect costs of the technology on which the environments run.

Can we advise you?

  Webinar Multicloud (1) carlos valverde

 

Carlos Valverde

 

Similar posts

Cloud migration with PaaS: optimize costs and business agility
8:01