Cloud TCO & AWS Hadoop Migration
Cloud TCO & AWS Hadoop Migration —
Specialty Retail Jeweler
2022  ·  Amazon Web Services  ·  Advisory & Migration
AWS TCO Analysis IaaS Pricing Like-for-Like Costing Hadoop Migration Oracle Databases Server Inventory Specialty Retail

I led a cloud TCO and migration planning engagement for one of the world's largest specialty retail jewelry companies. The engagement covered IaaS cloud pricing for their on-premises server estate, detailed like-for-like costing models comparing current datacenter spend against AWS equivalents, and a server-to-application mapping to validate migration scope and sequencing.

A key commercial outcome was an executed AWS Professional Services SOW for the Hadoop migration — migrating a significant big data workload to AWS. The engagement also covered Oracle database workloads in the RISC pricing model, which represented the highest-value servers in the estate by instance cost. The TCO model was iterated through multiple versions reflecting successive data quality improvements and stakeholder input.

~120
Servers Priced
Full RISC table pricing across the estate
~60
Cloud Cost Assets
CloudScape asset export used for rightsized pricing
12
TCO Model Versions
Spreadsheet iterated from v1.12 through final
SOW
Hadoop Migration — Executed
AWS Professional Services migration SOW signed

Executed AWS Hadoop Migration SOW

The primary commercial deliverable of this engagement was an executed AWS Professional Services SOW for the migration of the client's Hadoop big data platform to AWS. The Hadoop environment — used for retail analytics, customer behavior modeling, and inventory intelligence — was identified as a high-priority migration candidate due to the maturity of AWS EMR and S3 as Hadoop-native managed service replacements. I scoped the migration engagement, aligned on pricing and milestones with both the client and AWS, and drove the SOW to signature.

Like-for-Like Costing

  • Production like-for-like: current server specs directly mapped to equivalent AWS EC2 instances
  • Non-production like-for-like: separate non-prod environment sized and priced independently
  • Adjusted like-for-like: August 2022 version with refinements from client-provided corrections
  • Prod Costs workbook: production-only cost summary for stakeholder review
  • Oracle instances priced at BYOL on EC2 — highest per-instance cost in the estate

Rightsized Cloud Pricing

  • CloudScape cloud-cost-assets export provided rightsized instance recommendations per server
  • AWS US Virginia pricing applied (client's primary region) with reserved instance options
  • RISC table (~120 rows): full inventory with per-server AWS instance and hourly cost
  • Storage I/O costs included: EBS volume, S3-equivalent, and network egress estimated
  • Rightsized vs. like-for-like delta calculated to show optimization opportunity

Server-to-Application Mapping

Produced a server inventory with application association — linking physical and virtual servers to their owning application for migration wave planning and dependency validation.

Oracle Database Servers

Oracle database servers identified and priced separately in the RISC model. Oracle BYOL on EC2 r5a instances at highest per-server cost in the estate — licensing implications documented.

Error & Gap Resolution

Asset errors identified in May 2022 pass resolved in successive TCO iterations — ensuring clean data underpinned the final financial model and SOW pricing assumptions.

Informatica Workloads

Informatica ETL servers identified in the CloudScape assets — sizing and migration path noted for the data integration layer supporting retail analytics pipelines.

01

TCO Financial Modeling

Built and iterated the TCO model through 12 versions — covering like-for-like, adjusted, rightsized, and production-only scenarios for AWS across the full server estate.

02

IaaS Cloud Pricing

Produced the full RISC table pricing (~120 servers) with per-server AWS instance mapping, hourly costs, and storage I/O — the detailed cost foundation for the migration business case.

03

Hadoop Migration SOW

Scoped the Hadoop-to-AWS migration engagement, aligned pricing and milestones with the client and AWS, and drove the Professional Services SOW to signature.

04

Oracle Licensing Analysis

Assessed Oracle database licensing implications for AWS migration — BYOL on EC2, RDS Oracle licensing model, and the cost delta between options for the client's Oracle estate.

05

Application Mapping

Produced the server-to-application inventory linking servers to owning applications — enabling wave sequencing, dependency review, and migration scope validation with the client.

DeliverableDescriptionFormat
TCO Spreadsheet (v1.12–v1.23)12-version iterative TCO model — like-for-like, rightsized, adjusted, and production-only AWS pricingXLSX
IaaS Cloud Pricing — RISC TableFull ~120 server RISC pricing with AWS instance mapping, hourly costs, and storage I/OCSV + XLSX
Like-for-Like Costing (Prod)Production environment like-for-like cost comparison — current datacenter vs. AWSXLSX
Like-for-Like Costing (Non-Prod)Non-production environment separate costing — sized and priced independently for TCO accuracyXLSX
Cloud Cost Assets (Rightsized)CloudScape-generated rightsized asset export used as cross-validation for manual RISC pricingXLSX
Server Inventory — App MappingServer-to-application association workbook linking physical and virtual servers to owning applicationsXLSX
AWS Hadoop Migration SOWExecuted AWS Professional Services SOW for Hadoop platform migration to Amazon EMR / S3PDF Executed