I recently grabbed some notes about cloud migration planning on a course from cloud core concepts from David Linthicum. This section was very interesting to me and I decided to keep it as a reminder.
How to measure operational impact in a migration
To measure the operational impact we need to develop three cost models.
- Total cost of investment (TCO): all costs calculations from the migration including the Transactional costs.
- Transactional costs (TC): the added cost of the migration. Including what to do with leftover hardware and software and organizational transformation.
- Return of investment (ROI): the full program cost vs. value calculations.
Metrics to measure the migration impact
We need to understand what is important to data in the cloud, looking at value categories. Only then we can understand the impact of each category by providing a weight, or what it means to our enterprise.
- Improve agility
- Improve productivity
- Improve scalability
- Decrease fo cost
Assessing the operational impact
As part of the migration process, we need to assess the operational impact of these steps.
1. The current state assessment
This is basically how things are now, or the as-is.
We need to do an application inventory state-level assessment by doing:
- A sample applications assessment
- A full estate inventory
- An defining the application migration approach.
On the other hand, we need to do a service levels and operational environment assessment by doing:
- The current infrastructure.
- The operational costs.
- The current state architecture qualities and requirements.
- The current state operational models and processes.
2. The future state architecture
The to be, architecture or what the cloud deployment will look like once it's done.
3. The total cost of investment (TCO)
Develop the TCO model including all costs calculations from the migration, plus the Transactional costs.
4. The business benefits (ROI)
Finally, we can develop the ROI as well as create a budget
Cloud migration planning steps
- Replace: replace the software with a SaaS solution.
- Reuse: dividing up the application's assets and, leveraging those services for other applications.
- Refractor: Altering the application to be more cloud-native, and thus more efficient.
- Replatfom: placing the application on a different cloud platform.
- Rehost: lifting and shifting an app to the cloud with no modification.
- Retain: not moving the application to the cloud.
- Repair: remove the application completely from the service.
Requirements selecting the cloud provider
When selecting a cloud provider, helps to focus on the main parameters making a list of the needed requirements for the project. These requirements can be:
- Flexible resource configurations.
- Dynamic scale-up / scale-down resources.
- Seamless support of multiple clouds.
- Flexible resource quotas.
- Role-based controls.
- Comprehensive monitoring and logging.
- Image lifecycle management.
- Incident and change management.
- Service provisioning.
- End-to-End automation.
- Support APIs.
- Self-service resource provisioning.
- Rapid elasticity.
- Capacity on demand.
- Seamless support.
- Metering and chargeback.
- Pay-as-you-go consumption
- Asset tracking and usage reporting.