A Playbook For Effective Cloud Migration – Part 2 - Architecture & Governance Magazine

(Editor’s Note: Part 1 appeared yesterday here.)

Comparison of AWS, Microsoft Azure and Google Cloud Services

The following table describes the high-level comparison of Cloud Migration/adoption Services from AWS, Azure and Google Cloud.

Feature/Services	AWS	Microsoft Azure	Google Cloud
Application Migration	AWS Application Migration Service (MGN)	Azure Migrate	Google Cloud Migration
Database Migration	AWS Database Migration Service (DMS)	Azure Database Migration Service	Google Cloud Database Migration
Infrastructure Migration	AWS Server Migration Service (SMS)	Azure Site Recovery	Google Cloud Migration for Compute Engine
Application Discovery	AWS Application Discovery Service	Azure Migrate Discovery	Google Cloud Application Discovery Service
Key Features	Lift-and-shift, Modernization, Minimal Downtime	Lift-and-shift, Assessment, Modernization	Lift-and-shift, Minimal Downtime, Data Replication
App Hosting	Amazon elastic Beanstack	Azure Cloud Services	Google App Engine
On-Premises-to-Cloud Peering	AWS Direct Connect	Azure ExpressRoute	Cloud Interconnect
Compute Services	EC2, Lambda	Virtual machines	Compute engine
RDBMS Database Services	RDS	SQL Database	Google Cloud SQL
NoSQL: Key-Value	Amazon DynamoDB	Azure Cosmos DB	Google Cloud Datastore, Google Cloud Bigtable
NoSQL: Indexed	Amazon SimpleDB	Azure Table Storage	Google Cloud Datastore
Data Protection	AWS Key Management Service, AWS Certificate Manager	Key Vault, App Service Certificates	Cloud KMS & Cloud HSM
Storage Services	Amazon Simple Storage (S3), EBS	Blob Storage, Azure Files	Cloud Storage
PaaS	AWS Elastic Beanstalk	Azure App Service	Google App Engine
Configuration	AWS Config	Azure Policy and Azure Automation	Google Cloud Configuration Service
Container	AWS Elastic Container Service, Elastic Kubernetes Services	Azure Container Apps, Azure Kubernetes Service	Google Cloud Run, Google Kubernetes Engine
Serverless Functions	AWS Lambda	Azure Functions	Google Cloud Functions
Elastic Load Balancer	Elastic Load Balancing (ELB)	Azure Load Balancer	Cloud Load Balancing
Content Delivery Network	Amazon CloudFront	Azure CDN	Cloud CDN
Key Management Services	AWS KMS	Azure Key Vault	Google Cloud KMS
Supported Sources	On-premises, VMware, Other Clouds	On-premises, VMware, Hyper-V, Other Clouds	On-premises, VMware, Other Clouds
Schema Conversion	AWS Schema Conversion Tool (SCT)	Azure Schema Conversion Tool (SCT)	Google Cloud Schema Conversion Tool (SCT)
Identity & Access Management	AWS Identity & Access Management (IAM), IAM Identity Center, Amazon Cognito, AWS Directory Service	Microsoft Entra ID, Azure RBAC	Cloud Identity & Access Management, Identity Platform
Network & Application Protection	AWS Web Application Firewall and AWS Shield	Azure Web Application Firewall, DDoS Protection Service	Google Cloud Armor, Google Cloud Armor Managed Protection Plus
DNS Services	AWS Route S3	Azure Traffic Manager	Cloud DNS
Log Monitoring	Amazon Cloud Trail	Azure Operational Insights	Cloud Logging
Performance Monitoring	Amazon CloudWatch	Azure Application Insights	Stackdriver Monitoring
DevOps	AWS CodePipeline, AWS CodeBuild, AWS CodeDeploy, AWS CodeStar	Azure Boards, Azure Pipelines, Azure Repos, Azure Test Plans, Azure Artifact	Cloud Build, Artifact Registry, Tekton, Google Cloud Deploy, Binary Authorization, Operations Suite
AI & Machine Learning	Amazon Polly, Amazon Transcribe, Amazon Lex, Amazon SageMaker, Amazon Bedrock	Azure AI Immersive Reader, Azure AI Metrics Advisor, Azure Machine Learning, Azure Content Moderator	Vertex AI Studio, Translate AI, Gemini Code Assist, Document AI, Dialogflow
Automation	AWS Opsworks, Automated Migration Planning, AWS Migration Hub	Azure Automation, Automated Migration Planning, Azure Migration Center	Compute Engine Management, Automated Migration Planning, Google Cloud Migration Hub
Compliance	AWS Cloud HSM	Azure Trust Center	Google Cloud Platform Security
Monitoring	Amazon CloudWatch	Azure Monitor	Google Cloud Monitoring Service
Business Analytics	Amazon QuickSight, Amazon Kinosis	Azure Stream Analytics, Azure BI Tool	Cloud Dataflow, BigQuery
Hybrid & Multi Cloud Capabilities	AWS Direct Connect extend its cloud services into non-cloud environments, offering enhanced security and observability	Azure ExpressRoute connects the cloud to on-premises services	Google Anthos allows to run applications on-premises, in the cloud, or across multiple clouds, providing flexibility and avoiding vendor lock-in
Open-Source Support	WS CLI, SDK and CDK supports Open Source	Strong Open Source Support. Visual Studio Code	Strong Open Source Support. Knative, a serverless solution for running applications on Kubernetes

In summary, AWS Cloud Platform offers lot of tools and services for the cloud adoption, provides robust security features. The platform is highly scalable and reliable.

Microsoft Azure Cloud Platform offers vast cloud services, provides strong Microsoft integration features and supports hybrid cloud strongly. Azure is the best option for the existing Microsoft infrastructure.

Google Cloud Platform supports hybrid cloud environments and is best suited for next generation technologies like AI and ML. It provides high levels of Open-source technologies.

Best Practices for Cloud Migration

Cloud Migration is a sequence of steps to be performed to have cloud migration objectives of the organization met as per the business drivers. Based on the author’s experience on various Cloud Migration initiatives, 70% of businesses have only one application on the Cloud. There are several Cloud migration approaches and can be customized or fine-tuned based on the business drivers and available application inventory.

Below are the best practices that help to manage organization applications and solutions in the most secure manner on the Cloud.

A. Change in philosophy: in the cloud model best practices recommend is to “design for failure” instead of “design not to fail”, which is a totally different (but robust) approach

B. Perform POC: As part of the proof of concept, choose small application as a case for the migration. Define the migration plan and migrate the application. Assess the gaps and arrive at the master migration plan. Test the workload and estimate the resources like storage size, number of VM’s required, network bandwidth and security controls that an application requires for the migration. Design a framework to ensure that the cloud migration meets all the requirements and assures compatibilities for cloud journey. The POC helps the stakeholders to change their mindset and understand the benefits of cloud migration.

C. Move Group Identification: The identification of applications which can move on to the cloud are classified as Move Groups. Move Groups are logical grouping of applications, which transformed and migrated together to cloud as per the defined timelines. The move groups are identical to both production and non-production environments. The applications in scope are logically grouped together into Move Groups for transformation and migration within a defined timeline. More than one move group may run in parallel, and a move group may start after completion of other.

D. Application migration approaches: Need to adapt various migration approaches depending on short term and long-term business/technical goals. These approaches classified as 6R

Re-hosting: It can be done using automated tools or manually. It’s called “Lift-and-Shift” of the applications
Re-platforming: Determine the new platform and modify the underlying infrastructure. No change to the existing architecture
Replace: Moving to a different product. Its preferably, moving to SaaS platform
Re-architecting: Redesign the application and Infrastructure architecture using cloud-native features
Retire: Decommissioning of the application
Retain: Continue to use the application as is state

E. Multi Cloud Environment: This step helps to decide on the application to run on single cloud environment or multiple cloud environment. Single cloud provider leads to vendor lock in. Moving the application to a different cloud provider consumes more effort. The various models for using multiple cloud providers are,

One application in one cloud: One category of applications runs in one cloud provider and another set in another. This approach gives increased business advantages and flexibility.
Split application across multiple cloud providers: Parts of an application run in one cloud provider and other parts of it in another. This approach helps to leverage the advantages of each cloud provider.
Cloud agnostic Application: These types of applications run on any cloud provider. The application simultaneously runs on multiple providers or split application load across providers. This model gives the flexibility to shift loads from one cloud provider to another

F. Automation: Automation provides the ability to build infrastructure as code and automatically deploy the applications without any downtime. During the migration of multiple applications, organizations come across repeatable patterns that need to be automated. It will cut down on migration time, provide more consistency, and spread an automation thought process across the teams so that they can see the real benefit of moving to the cloud. Adopt migration factory that leverages remote delivery for scheduling and monitoring of jobs. This reduces cost of labor, time and onsite professional services for cloud journey, drastically.

G. Data Migration: Conduct a data assessment of the candidate identified and complete data migration from production to Cloud. Organizations sent the on-premises data sets to the cloud provider to be uploaded into the cloud. Still host application in the existing customer data center and point it to the data migrated in cloud. This is usually the best method to migrate large volumes of data in bulk.

H. Virtualization: This is quick and easy migration to cloud as no changes will be required to the application. These workloads transferred to the cloud, by a mechanism called lift-and-shift or re-hosting process. This is the fastest way to migrate applications, but the process requires proper instance sizing and configuration optimization to meet performance and user experience goals. This is suitable for legacy applications.

I. Database migration: This process covers data replication, data loading and moving multifaceted databases to cloud databases. Plan for proper archival and backup strategy for data. As a best practice, keep the dynamic data close to compute and static data close to the user. This can be done by leveraging traditional caching techniques.

J. Integration: Cloud migrated applications shall communicate over internet to the applications on premise. This could lead to performance and bandwidth-related issues. Need to develop high performance applications.

K. Licensing: Check for current set of tools/software used to support cloud-based licensing model. For ex: planning to leverage the elasticity of cloud computing make sure that licenses are compatible

L. Vendor lock-in: Do not utilize the cloud vendor services, which leads to lock-in with vendor. Need to weigh in the pros and cons of which services to use and what extent.

M. Budget Planning: The Cloud computing survey report goes on to highlight that businesses are likely to spend 28% of their IT budgets on building a better Cloud infrastructure for their organization. A move to the Cloud can influence the CFO in two ways – A better cash flow and Capex converting to Opex. It is up to the organization to prioritize and decide whether this is in line with the organization’s financial goals.

N. Monitoring and Governance: Build a Cloud Center of Excellence with multiple cross-functional teams of people who are responsible for development & management of cloud strategy, governance and best practices for the organization. Organizations should adopt the agile methodology and work in sprints to perform mass migrations in addition to providing continuous learning and improvement to the entire migration process.

O. Migration Architect: Establish the migration architect role to lead the effort. The migration architect is a system architect-level position responsible for planning and completing all aspects of the migration. The core responsibility of Migration Architect include defining necessary refactoring required to make the migration successful, designing strategies for data migration, defining cloud-solution requirements, and determining migration priorities and production switchover mechanisms. This will eventually leverage the rest of the business units in organization to shift to cloud technology, smoothly.

P.Skill Management: The end goal for the team is the effective management of resources to keep applications up and running and meet business goals. The team consisting of Operations, development and design needs to be trained on cloud and its design principles. The team also to be trained in migration to develop a template and automation around the migration.

Summary

A move to the Cloud is inevitable. Migrating organization applications to cloud results in reduction or eliminate costs, accelerate time to market, and drive additional revenue, freeing up resources to innovate and focus on core business. In summary, the following are the recommendations from the author based on various customer engagements performed in Cloud Migration space,

Choose the right Cloud provider, each provider has specific strengths and weaknesses. Evaluate them properly for the fitment of the customer needs
Usage of right migration tools. No Vendor Lock-In of the application
Move to agile operating model
Mostly, Private and Hybrid Clouds to replace data centers of the customers
Huge demand from customers on reducing Capex and Opex
Maximize productivity with scalability and high availability
Always migrate Development and Testing environments on to cloud
Always, move Core functions to private cloud and non-core to public cloud
Wherever possible, adopt hybrid model, its safe bet
Transform architecture to a scale out capability and enabled for dynamic API based interaction
Bring cloud native capabilities to application and transform to deliver real-time capabilities and work on actionable data
Drive cost takeout by driving higher level of automation and reducing projects and run costs
Having skilled staff ready. Staff need to know all the new technologies and processes. For ex: Microservices, DevOps, API Management, AI etc.

Additionally, there is no one-size-fits-all answer when choosing between AWS, Azure and GCP. The best choice always depends on the specific use case. GCP excels in AI, machine learning, and big data applications. Azure is particularly strong in integrating with Microsoft infrastructure. AWS, however, stands out by offering the most diverse range of services, making it ideal for complex requirements and large-scale projects.

CXO’s, Architects and IT Leaders can leverage the structured Cloud Migration Framework explained in this playbook that ensures the alignment of Organization goals and objectives. Also, the Leaders can adopt the recommendations from this playbook to suit their organizations needs from simple Lift and Shift to Re-architect and hybrid cloud solutions.

References

1.https://www.opensourceforu.com/2024/11/managing-a-hybrid-cloud-an-overview/

Acknowledgements

The authors would like to thank Tanay Srivastava, Director, Tricon Solution LLC for giving the required time and support in many ways in bringing up this playbook as part of Technical Services efforts.

About the Author

Dr. Gopala Krishna Behara is an Enterprise Architect at Tricon IT Solutions. He has around 28 years of IT experience. He can be reached at gopalakrishna.behara@triconitsolutions.com.

Disclaimer

The views expressed in this article/presentation are that of authors and Tricon Solutions and does not subscribe to the substance, veracity or truthfulness of the said opinion.