SKT DataLake workload integration

Background

As AI and data-driven decisions become central to competitiveness, industries are moving from on-premises systems to scalable, flexible cloud data platforms. A prerequisite is the systematic classification and migration of diverse workloads, with standardized validation to preserve quality and business continuity after transition.

SKT launched the DataLake Workload Integration project to consolidate distributed on-premises platforms on AWS Cloud, systematically migrate large workloads and establish stable operations.

Scope

• Workload analysis and migration
• Data security and validation framework
• Automated validation processes
• Analyst workload migration support

Implementation

1. Workload analysis and migration:

We closely analyzed user environments and existing business logic, planned migration by job type and data flow, and developed programs suited to AWS Cloud DataLake. This enabled a stable transition from on-premises processing to the cloud.

2. Data security and validation framework:

We established and applied security controls and a standard framework to compare and validate large datasets between production and migration environments, verifying consistency and accuracy.

3. Automated validation processes:

We automated collection of validation targets and reference information and centralized results, creating a sustainable process that consistently validates new data processing jobs.

4. Analyst workload migration support:

We analyzed analysts' work environments in advance, identified migration targets and job dependencies, and supported systematic planning and stable operations.

Results and expected benefits

Consolidating distributed on-premises platforms on AWS Cloud established a stable, scalable foundation for data processing.

Systematic job-by-job migration and precise validation minimized transition errors and improved stability. Standardized, continuous checks secured data consistency, quality, reliability and business continuity. Automated validation established enterprise quality management that maintains consistent standards after migration.

The preliminary data-flow analysis provided a basis for systematic migration schedules, helping analytical work settle smoothly and reliably into the cloud.