Migrating to the clouds:

Strategy, Flight plan and checklist

Introduction

Gartner reports that by 2025, 51% of IT spending in application software, business process services, and infrastructure will be on cloud platforms. Organisations embrace this radical shift to leverage new infrastructure, business processes, or the cloud’s latest technologies. Many of them have already accelerated this movement due to the pandemic in the last two years. Data migration or refactoring is invariably required when on-premise applications are moved to the cloud. Thus, there is an increasing focus on data re-engineering necessitated by cloud migration.

The movement to the cloud initiates a “migration”, “modernisation”, or both. These two terms imply slightly different approaches. In the former case, the source and destination architecture remain the same; in the latter case, there is some transformation, big or small, to the existing application architecture and data schema before they are migrated. Application and data migration is relatively straightforward, whereas modernisation needs to be executed after considerable assessment and planning.

We accomplished a seamless data modernisation for a client using an eight-point approach. We generalise the methodology into a framework that other projects can orchestrate for successful implementation given as:

1. Migration strategy definition
2. Data analysis
3. Migration script development (data and file)
4. Meta-data driven file migration
5. The Agile team, Scrum process and roles
6. Agile data migration (iterative)
7. Testing and validation
8. The cut-over plan

Why Modernize

Each organisation may have its own rationale for modernisation; however, the following highlights the major objectives that triggered modernisation and cloud migration with our clients:

• The on-premise infrastructure from a third party is nearing its end of life and needs decommissioning.
• The decade-old legacy application has accumulated technical debt and performed poorly with the increase in users and trans actions.
• Historically, the users had access to the application file repository. Manual file naming disregarded naming conventions, making the files and their mapping in the SQL table out-of-sync with each other.
• Diminishing data quality due to many patch fixes and ad-hoc modifications done on the application for more than a decade. This is also an opportunity to improve data quality
• The legacy infrastructure does not scale up to meet the compute or bandwidth needs resulting in poor user experience.
• Software licenses nearing their end of support phase; this is the right time to move into a modern PaaS framework that provides optimum upgrades at no upfront cost.
• By moving the on-premise files to Azure, cheaper, reliable storage is available on a scalable infrastructure. Also, the cold storage options on Azure bring greater incentives in managing high volume archives.
• Leverage gains in performance, licence cost, patch management etc., by moving from the SQL server 2008 to the Azure SQL DB.
  

Pages: 1 2 3 4