CLOUD-NATIVE MIGRATION SERVICES
UNDERSTANDING CAPABILITIES & LIMITATIONS
Introduction
Transitioning from the traditional Datacenters for migrating the workload to the cloud is one of the most crucial and difficult decisions for any business and organization. Unfortunately, most organizations struggle along this transition journey due to inertia for change within an organization, external pressure due to competition, and legal compliances. Other prominent factors also include challenges because of the fear of Cloud adaptation to enterprise architecture and the inability of the tech community to deliver with agility and speed.
Concerns specific to the industry domain, regulations, and organizational inertia require management and leadership interventions to overcome.
On the other hand, challenges due to technology adaptation can be minimized/avoided, to a certain extent, through the appropriate selection of technology, tools, process, and operations suite.
This brings us to an important technology aspect — understanding that the organization is now ready for data center migration and has its Cloud Adoption Framework defined and established. So, what is the choice of tooling available to support? Especially when migration to the cloud has different propositions which are defined by various migration R’s. This brings us to unleash the capabilities and leverage the Cloud Native Migration Services.
Cloud-Native Migration Services
Every major cloud provider, viz-a-viz, AWS, Microsoft Azure & Google Cloud offers the suite of services, tools, frameworks, guidelines, and best practices to navigate smoothly through their cloud migration journey. These services are bundled under AWS Migration Service, Azure Migration Service, and Migrate for Anthos by GCP. Solutions for the workload migration around these suites of services and tooling are recommended nowadays in most solution plans. The brief overview (non-exhaustive) of the services available across public cloud providers is depicted as below:
Let’s discuss specific characteristics and offering of these services:
Cloud-Native Advantage
The tools and services natively support different integrations with their respective cloud ecosystem, which encompasses application, database, networking, governance, and telemetry. It is a clear advantage over bespoke or in-house migration accelerators. The integration is mostly vetted against optimization, security & speed required for successful cloud migration. This scenario is so promising that we can migrate any legacy application as a container workload to any managed container service today. For example, containerize legacy .Net application and deploy on containerizing web application hosted on Azure App Service or Google Cloud Run.
Supporting Disposition R’s
Support for various R’s of cloud disposition is fundamental for any cloud migration. These cloud-native migration tools and utilities allow us to deploy various kinds of workloads to the cloud. This workload transformation is majorly supported for Rehost, Refactor, or Replatform propositions. With Azure Migrate Server Assessment and Azure Migrate Server Migration, it is safe to do as-is lift-and-shift of the workloads to VMs. A similar kind of tooling support is available from AWS & GCP for rehosting & re-platform the workloads to migrate them to containers, PaaS web (e.g, Azure Web Apps). It supports applications, transformation, and modernization for legacy, critical and high-volume databases.
Conformance to Application Migration Lifecycle
The matrix clearly shows the choice of tooling and options available to support each phase of application migration. Except for some gaps, the coverage seems very robust.
Technical Agility
These tools & services are consistently UI-driven with sufficient documentation and minimal learning requirement. As a result, cloud engineers can easily adopt them to deliver migrated workload in no time. Augmenting to their cloud tools arsenal: they are equipped to deliver migrated workload of any kind — databases, middleware, custom, or COTS applications
Automation Acceleration
Automation is a key aspect of accelerating workload migration. Once the migration discovery, assessment, and migration jobs are configured we only need to monitor the migration progress until its completion.
Varied Architecture Support
Infrastructure support computes adaptation for platform services, and the ability to support serverless and serverless disposition. Wide variety of application and database architecture supported: this includes standalone monoliths, multi-server 3-tier farms, distributed workloads specifically databases.
Roadmap & Vendor Support
Vendor support and the future roadmap of having advanced capabilities to support workload migration is the key aspect available with these cloud-native migration services. Migrate for Anthos or Azure Migrate all are managed services with production support SLAs. This provides much-required confidence during migration.
Conformance to Well-Architect Framework
Security adherence with cloud ecosystem, performance efficiency with inherent automation, support for site reliability, and SLA, reducing the cost of migration by enabling operations & outcome measurability are the benefits of using respective cloud-native services. Thus, this conforms to the Well Architect framework adoption and requirements.
Migration Validation
Data migration services such as Azure DMS (Data Migration Services) can support the heavy lifting of database workload and provide mechanisms to generate metrics, migration reports, sanity testing, schema validation, and data-type mapping. Thereby providing a comprehensive suite for data migration.
Operations & Controls
Cloud-enabled telemetry support integrated with the services such as Google Cloud Monitoring with Google Anthos allows capturing metrics during migration. This helps debug during the migration process or to track any adverse behavior. Inherent controls, security, IAM measures provide the required security and access monitoring.
Cost Advantage
The cloud-native proposition is leverage and enabling a decision to adopt for the native services. It is a comparatively efficient approach instead of bespoke or third-party vendor CoTS. Enabling and integrating with other cloud-native services is the inherent and overall value proposition.
Performance Baseline
Estimating and right-sizing the target environment is challenging for any workload migration. We tend to match source infra to similar destination infra with a lift-and-shift approach. For the small workload, this seems to be an efficient option, but this is most inefficient for the multi-server interleaved environment. The alternate approach is to estimate the right size based on performance metrics captured over a significant duration of time. This approach, though time-consuming but an efficient way to control the cost.
Cutover & Rollback
Application /database migration requires various cutover strategies depending upon the business criticality and SLAs. These native tool offerings support scenarios such as lift & shift, then cutover, or replication-based parallel runs alongside the source application/database and canary deployments. Similarly, support for the previous state’s rollback can be configured at certain levels.
Conclusion
The cloud-native migration services by AWS, Azure, and Google Cloud offer the non-exhaustive tools and utilities to plan, prepare, execute, and optimize the workload migration to the cloud. These services are capable of accelerating the migration journey for some typical workloads, such as rehost, replatform, and refactor. However, the other use cases require major overhauling and rearchitecture to modernize the specific workload and application. Certainly, there are deficiencies and limitations with the tool stack and utilities. Having said that, the roadmap and vendor support ecosystem is promising to confidently transition the applications and workloads to the cloud. Nevertheless, the described features’ benefits accelerate the workload migration process by streamlining and performing heavy lifting. Thus, adapting to the Cloud Native migration services is a go-to choice that should be incorporated into any workload migration strategy.
Note: This article was first published by the author on Feb 03, 2022, on Mphasis website’s Thought Leadership’s blog available at this URL
