1) Lift-and-Shift Migration
The lift-and-shift migration targeted over 8000 VMs located all over the world, in small datacenters. So it's safe to say that the number of the servers targeted for migration was big and that it became imperative to plan effectively the migration journey.
After deciding on a migration approach, it was primordial to assess the on-premises apps, databases and infrastructure. With the help of automated cloud migration tools, insights into dependencies were extracted and an inventory regarding infrastructure was made. This way, we were able to get right-sizing guidance, workload-level cost estimates, and performance metrics.
Next step was to account for workload priorities, timelines, milestones, resources, and funding. The overall plan was then broken up your into smaller migration projects, each with a group of related workloads.
To avoid issues during and after migration, by using best practices, we pre-configured landing zones regarding networking, identity, management, security, and governance elements that took into consideration all the internal organizational standards.
Finally, as migration sub-projects were finished, securing the environment became the primary focus. By benefiting from Azure Security Center and setting up cloud-native alerts and SIEM solutions, the environment now quickly detects and responds to threats with intelligent threat protection.
2) Abstract platform for automated deployment
Once migrated to the cloud, the organization quickly understood that true value lies within PaaS services. For new projects, especially the ones oriented to be cloud-native, these types of services needed to be available for any member of the IT department who needed infrastructure for new projects. As most of the members were lacking cloud skills, it was needed a way to abstract all this and "translate" cloud terminologies into normal IT language.
With the help of a ticketing solution and ARM templates, the organization managed to overcome this issue. Now, all IT members execute the following process when new IT resources are needed for new projects:
1) Submit new ticket in the ticketing solution - the ticket has a predefined template which requires details about the services needed and parameter values;
2) Once the ticket is submitted, it goes through an approval process;
3) Once the ticket is approved, it automatically triggers pipelines in Azure DevOps that executes ARM templates. These ARM templates are responsible for creating the exact resources that the IT member requested;
4) Once the services in Azure are created, an automated email is sent to the creator of the ticket with details about how to connect to the infrastructure/services newly created;
To pull off such automation, a complex architecture of orchestrations and pipelines has been developed in Azure DevOps. Also, over 200 ARM templates were created for different infrastructure scenarios, serving as building-blocks for any project need that might occur in the future.
