Hybrid Cloud Mesh

Hybrid Cloud Mesh

Join this online group to communicate across IBM product users and experts by sharing advice and best practices with peers and staying up to date regarding product enhancements.

 View Only

Migrating Applications from AWS to Azure - The hard way or the easy way!

By Dave Hegenbarth posted Tue March 25, 2025 05:36 PM

  

 



DON’T DO THIS – IT TAKES TIME 

Migrating Applications from AWS to Azure with Apptio, Turbonomic, Terraform, and Manual Networking Configuration 

When migrating an application from AWS to Azure, where both clouds are already set up, Apptio and Turbonomic analyze cost and performance data to determine optimal workload placement. Terraform automates infrastructure provisioning and the deployment of computing, storage, and networking resources in Azure. However, manual network configuration is required to ensure secure connectivity between AWS and Azure and to handle an application’s upstream and downstream dependencies. To execute the migration, the networking team must: 

  • Prepare the migration checklist/discovery: 

  1. Which applications are moving? And how many in each window? 

  1. What dependencies does each have, and what IP addresses and ports do they need to reach for those dependencies? Likewise, the return path needs discovery. 

  1. If the documentation doesn’t exist, it needs to be created. If it does exist, it needs to be validated and updated if necessary. 

  • Plan the migration/submit the change requests: 

  1. Routing. Evaluate and potentially update VPC Peering, Azure VNET Peering, VPNs, ExpressRoute, & Direct Connect. 

  1. Security. Update security groups, firewall rules, etc. 

  1. DNS. Which private and public records must be modified to consider the IP address changes? How long do those changes need to converge globally? 

  • Plan and prepare for downtime: 

  1. How many windows will be needed to complete the entire migration? 

  1. Which applications and services will be impacted, and who will they affect? 

  1. How much notice is required? Are there SLA’s that need to be adhered to? 

  1. Contingency: Plan for the worst, hope for the best. What is the post-migration test plan, and how much time is needed to complete the tests? How much time is needed in the window for a rollback?  

  • Post migration, normalization: 

  1. Load balancers and traffic management rules must be adjusted to shift workloads gradually while maintaining availability. 

Meanwhile, the Application teams need to review the application code and, where necessary, reconfigure RPC/gRPC and API calls. In some cases, they may also change hardcoded IP addresses to reflect changes in the infrastructure and network routing.  

This approach provides full control but can introduce significant delays due to network complexity, team coordination, and migration downtime. 

-------------- 

DO THIS – IT SAVES TIME 

Migrating Applications from AWS to Azure Using Hybrid Cloud Mesh with Terraform 

With Hybrid Cloud Mesh and Terraform, migrating an application from AWS to Azure becomes significantly simpler and faster, eliminating the need for complex manual networking changes.  

Apptio and Turbonomic identify cost-saving opportunities and Terraform provisions the required Azure infrastructure. Hybrid Cloud Mesh abstracts the application connectivity requirements from the underlay network, decoupling the application/service name from its’ associated IP address maintaining the service continuity throughout the migration, which not just eliminates the downtime and need for planning for that downtime and it can also eliminate and/or hugely simplify the networking configuration: 

  • Prepare the migration checklist/discovery: 

  1. What apps/services do you want to move, and where would you like to run them? Hybrid Cloud Mesh maintains its upstream and downstream dependencies, allowing for progressive migration with zero service interruption.  

  • Plan the migration/submit the change requests: 

  1. No downtime, no need to change the underlying network configuration and possibly no need to change firewall or security groups. So more than likely no change requests. 

  1. Instead of configuring VPNs, VNET Peering, or DNS updates, Hybrid Cloud Mesh creates an application-layer virtual overlay network that maintains service-to-service connectivity across clouds using existing infrastructure. 

  • Plan and prepare for downtime: 

  1. No downtime, no notice required, no need for weekend work. It can even allow for the Hybrid Cloud Mesh administrators to put guardrails in place to allow self-service connectivity for the Application teams. 

  • Post migration, normalization: 

  1. With Hybrid Cloud Mesh traffic shifting can be done gradually without breaking API calls. Allowing you to create active load-balancing or DR application architecture between Azure and AWS. Or make the cost save and turn off and remove the old deployment whenever you like.  

  1. And since the app connectivity is now decoupled from the underlay IP network, you can now also review and tune your underlay IP network for better performance, security or resiliency without worrying about impacting on the application traffic.  

This reduces networking dependencies, accelerates migration, and prevents downtime, allowing application teams to execute cloud migrations with minimal networking intervention. 

 

For more information go to - https://www.ibm.com/products/hybrid-cloud-mesh 


#TechnicalBlog
#BestPractices

0 comments
7 views

Permalink