Re-platforming to Cloud

Transform legacy systems to modern x86 and cloud platforms, reducing operational costs by 40-70% while enabling innovation and agility.

Service Overview

Today's preferred target for most workloads is x86—running Linux or Windows—enabling organizations to leverage both virtualization and cloud hosting. arqitekta specializes in transforming legacy systems to these modern platforms, making full use of virtualization and both private and public cloud environments.

Frequently, applications can be re-platformed with minimal effort because they rely on middleware and databases that function identically on x86 architectures. The key is to test the application on newer middleware versions or secure ISV support for the migration.

For workloads tightly coupled to legacy operating system APIs, the challenge is greater, but not insurmountable. We guide you through the process, and where necessary, connect you to specialist code refactoring partners who can transform source code from legacy platforms to x86. Throughout, our focus is on maintaining business continuity while unlocking the flexibility and efficiency of cloud-ready infrastructure.

Organizations that delay re-platforming face compounding risks: rising maintenance costs on end-of-life hardware, shrinking talent pools for legacy skills, and an ever-widening gap between what their infrastructure can deliver and what the business demands. arqitekta's approach mitigates these risks through a structured, evidence-based methodology that has delivered successful migrations across industries and platform generations.

The Re-platforming Imperative

Why Re-platform Now

Legacy Cost Escalation

Annual Cost Trends (per MIPS / per core):
- Mainframe licence renewals: 5-8% annual increases
- Legacy Unix support contracts: 10-15% above market
- Proprietary hardware refresh cycles: 3-5 year forced upgrades
- Specialist skills premium: 30-50% above x86 equivalents

Hidden Costs:
- Opportunity cost of frozen budgets
- Security patching delays on unsupported OS versions
- Compliance risk from aging infrastructure
- Integration friction with modern SaaS and APIs

Market Forces

Talent scarcity: COBOL, PL/I, and proprietary Unix skills are retiring faster than they are replaced
Vendor end-of-life: IBM Power roadmap consolidation, Oracle SPARC sunset, HPE Integrity discontinuation
Cloud economics: Reserved instance and spot pricing deliver 60-80% savings versus on-premises proprietary hardware
Regulatory pressure: Auditors increasingly flag unsupported platforms as material risk

Digital Transformation Dependency

Cloud-native CI/CD pipelines require x86 or ARM targets
AI/ML workloads demand GPU-enabled infrastructure unavailable on legacy platforms
Event-driven architectures and microservices assume container orchestration
Real-time analytics depend on elastic compute and modern data services

Common Re-platforming Mistakes

Boil-the-Ocean Scope

Attempting to migrate everything in a single wave
No prioritization by business value or technical risk
Ignoring the 20% of applications that drive 80% of complexity

Underestimating Data Gravity

Assuming data migrates as easily as compute
Ignoring EBCDIC-to-ASCII conversion edge cases
Neglecting referential integrity across phased migrations

Skipping the Proof of Concept

Moving directly from assessment to full migration
No baseline performance metrics for comparison
Insufficient validation of middleware compatibility

Our Re-platforming Framework

Phase 1: Discovery & Baseline

Weeks 1-3: Establish the Migration Foundation

Application Portfolio Analysis

Discovery Dimensions:
- Application inventory and ownership mapping
- Runtime dependency graph (middleware, libraries, OS calls)
- Data flow and integration point catalogue
- Performance baselines (CPU, memory, I/O, latency)
- Licence entitlement and contract review

Business Value Scoring

Revenue impact classification (critical, important, supportive)
Regulatory and compliance dependency mapping
User base and availability SLA requirements
Technical debt and maintenance cost per application

Platform Fit Assessment

For each application, we score against target platform options:

Cloud-native readiness: containerization feasibility, statelessness
Middleware portability: equivalent services on x86/cloud
Data migration complexity: volume, format, referential constraints
Integration coupling: upstream and downstream system dependencies

Phase 2: Architecture & Proof of Concept

Weeks 4-7: Validate the Migration Path

Target Architecture Design

Landing zone design (VPC topology, network segmentation, identity federation)
Compute sizing models (MIPS-to-vCPU, RISC-to-x86 conversion ratios)
Storage tier selection (block, object, file — performance vs. cost)
Database target selection with schema compatibility analysis

Proof of Concept Execution

Select 2-3 representative applications spanning complexity tiers
Stand up target environment with infrastructure-as-code
Execute migration, including data conversion and middleware swap
Run functional regression, performance benchmarks, and failover tests

PoC Success Gate

Gate Criteria:
- Functional parity: 100% test case pass rate
- Performance: Within 10% of baseline or better
- Data integrity: Zero data loss, format validation passed
- Rollback: Demonstrated clean rollback in &lt;30 minutes
- Cost model: Confirmed target TCO within business case range

Phase 3: Migration Factory

Weeks 8-20: Execute at Scale

Wave Planning

Group applications into 4-8 migration waves based on dependency clusters
Sequence waves to deliver early business value while managing risk
Allocate dedicated migration squads per wave (2-4 engineers each)
Define wave-level go/no-go criteria and rollback triggers

Standardized Migration Runbooks

Each wave follows a repeatable runbook:

Pre-migration: Environment provisioning, data replication setup, smoke tests
Migration window: Application cutover, DNS/load-balancer switch, data sync finalization
Post-migration: Functional validation, performance verification, monitoring activation
Hypercare: 5-10 business days of intensive support and incident response

Automation & Tooling

Infrastructure provisioning via Terraform or Pulumi
Configuration management with Ansible or Chef
Data replication using AWS DMS, Azure Migrate, or Striim
Continuous integration pipelines for migration validation scripts

Phase 4: Optimization & Modernization

Weeks 21-26: Maximize Value on the New Platform

Cost Optimization

Right-size compute instances based on actual post-migration utilization
Implement auto-scaling policies for variable workloads
Convert on-demand instances to reserved or savings plans
Decommission legacy infrastructure and terminate support contracts

Operational Excellence

Implement cloud-native monitoring (CloudWatch, Azure Monitor, Datadog)
Establish SRE practices: SLIs, SLOs, error budgets
Automate patching, backup, and disaster recovery
Integrate with enterprise ITSM and incident management

Modernization Roadmap

Identify candidates for containerization and Kubernetes migration
Evaluate serverless opportunities for batch and event-driven workloads
Plan API gateway deployment for legacy integration modernization
Sequence future modernization sprints based on business priority

Migration Patterns

Pattern 1: Lift and Shift (Rehost)

When to Use: Stable applications with urgent datacenter exit timelines Approach: Minimal application changes; move the workload to equivalent cloud infrastructure Benefits: Fastest time-to-migrate, lowest project risk, immediate infrastructure cost savings Limitations: Does not address technical debt or unlock cloud-native capabilities Typical Savings: 25-40% operational cost reduction Timeline: 2-6 weeks per application

Pattern 2: Re-platform (Lift and Reshape)

When to Use: Applications that benefit from managed services without full re-architecture Approach: Replace underlying infrastructure components — e.g., swap self-managed databases for RDS/Aurora, move from VM-hosted middleware to managed container services Benefits: Reduced operational burden, improved resilience, moderate modernization Typical Savings: 40-55% operational cost reduction Timeline: 4-10 weeks per application

Pattern 3: Re-architect (Refactor)

When to Use: Strategic applications with significant technical debt or scalability constraints Approach: Decompose monoliths into microservices, adopt cloud-native patterns (event sourcing, CQRS), rewrite tightly coupled components Benefits: Maximum agility, elastic scalability, full cloud-native capability Investment: 3-5x higher than rehost, but delivers the greatest long-term value Timeline: 3-9 months per application

Pattern 4: Retire or Replace

When to Use: Applications with low business value, redundant functionality, or available SaaS alternatives Approach: Decommission and redirect users to existing or commercial alternatives Benefits: Eliminate maintenance cost entirely, reduce portfolio complexity Consideration: Requires thorough data archival and regulatory retention compliance

Choosing the Right Pattern

Decision Matrix:
                    Low Strategic Value    High Strategic Value
Low Complexity      Retire / Replace       Re-platform
High Complexity     Lift and Shift         Re-architect

Technology Expertise

Source Platforms

Mainframe: IBM z/OS, z/VSE, AS/400 (IBM i), OpenVMS, Unisys ClearPath
Unix/RISC: IBM AIX (POWER), HP-UX (Itanium), Oracle Solaris (SPARC), Tru64 (Alpha)
Proprietary: HPE NonStop (Tandem), Stratus ftServer, Bull GCOS
Legacy x86: Windows Server 2008/2012, RHEL 5/6, CentOS end-of-life releases

Target Platforms

Operating Systems: RHEL 8/9, Ubuntu 22.04/24.04 LTS, Windows Server 2022, Amazon Linux 2023
Container Orchestration: Docker, Kubernetes (EKS, AKS, GKE), Red Hat OpenShift, Rancher
Public Cloud: AWS, Microsoft Azure, Google Cloud Platform, Oracle Cloud Infrastructure
Private/Hybrid Cloud: VMware vSphere/Tanzu, Nutanix AHV, Red Hat OpenStack, Azure Stack HCI

Migration Tools & Services

Discovery & Assessment: AWS Migration Hub, Azure Migrate, Flexera One, Device42, Cloudamize
Server Migration: AWS Application Migration Service, Azure Migrate, Google Migrate for Compute Engine
Database Migration: AWS DMS, Azure Database Migration Service, Ora2Pg, Striim, Ispirer
Data Transfer: AWS Snowball, Azure Data Box, Google Transfer Appliance (for petabyte-scale moves)
Code Conversion: Micro Focus Enterprise Developer, Astadia, LzLabs Software Defined Mainframe

Middleware & Database Mappings

Application Server Transitions:
- IBM WebSphere → Apache Tomcat, Red Hat JBoss EAP, Spring Boot
- Oracle WebLogic → WildFly, Quarkus, containerized Spring
- CICS/IMS → Spring Boot microservices, MuleSoft API layer
- BEA Tuxedo → gRPC services, event-driven architecture

Database Transitions:
- IBM DB2 → PostgreSQL, Amazon Aurora, Azure Database for PostgreSQL
- Oracle Database → PostgreSQL, Amazon RDS, Azure SQL
- Sybase ASE → PostgreSQL, SQL Server
- VSAM/ISAM → PostgreSQL, DynamoDB, Redis (for key-value patterns)

Messaging Transitions:
- IBM MQ → Amazon MQ, RabbitMQ, Apache Kafka, Azure Service Bus
- TIBCO EMS → Apache Kafka, Amazon SNS/SQS
- Proprietary queues → CloudEvents, Apache Pulsar

Industry Applications

Financial Services

Core Banking & Capital Markets Infrastructure

Typical Challenges

Mainframe-hosted core banking with decades of accumulated COBOL
Regulatory requirements for data residency and encryption at rest
Zero-downtime tolerance for payment processing and trading systems
Complex batch windows for end-of-day settlement

arqitekta Approach

Phase migration around batch processing windows to maintain settlement integrity
Deploy dual-run reconciliation: legacy and target systems process in parallel until parity confirmed
Implement hardware security modules (HSMs) and encryption-in-transit on cloud landing zones
Engage regulatory liaison early to pre-approve target architecture controls

Outcomes Achieved

55-65% reduction in mainframe MIPS costs
Batch processing time reduced from 8 hours to under 90 minutes
Achieved PCI-DSS and SOC 2 Type II compliance on target platform within 6 months

Healthcare

Clinical and Administrative System Modernization

Typical Challenges

Legacy EMR/EHR systems on proprietary Unix or minicomputer platforms
HIPAA compliance and PHI data handling during migration
HL7 and FHIR interoperability requirements across migrated and non-migrated systems
24/7 availability mandates for clinical decision support

arqitekta Approach

Implement encrypted data pipelines with PHI tokenization during migration
Use blue-green deployment patterns to ensure zero clinical downtime
Maintain HL7v2 interface engines while introducing FHIR APIs on the new platform
Conduct clinical workflow validation with practitioner sign-off at each wave

Outcomes Achieved

45-55% infrastructure cost reduction
EMR response times improved by 40% on cloud-hosted x86
Enabled telehealth API integrations previously blocked by legacy platform constraints

Manufacturing

OT/IT Convergence and Plant System Migration

Typical Challenges

SCADA and MES systems running on end-of-life Unix or Windows variants
Real-time process control latency requirements (<10ms response)
Air-gapped or semi-connected plant networks complicating cloud connectivity
Long certification cycles for changes to production control systems

arqitekta Approach

Deploy hybrid architecture: edge compute for real-time control, cloud for analytics and reporting
Migrate non-real-time workloads (ERP integration, quality reporting, planning) to cloud first
Implement OPC-UA gateways to bridge legacy SCADA protocols to modern API layers
Coordinate migration windows with planned maintenance shutdowns

Outcomes Achieved

40-50% reduction in plant IT infrastructure costs
Enabled predictive maintenance analytics on cloud-hosted data lake
Consolidated 12 regional plant servers to a single managed Kubernetes cluster

Retail & E-Commerce

Omnichannel Platform Modernization

Typical Challenges

Monolithic POS and inventory systems on aging x86 or midrange platforms
Seasonal demand spikes requiring 10-50x compute elasticity
Real-time inventory visibility across warehouse, store, and online channels
PCI-DSS scope reduction objectives

arqitekta Approach

Re-platform POS backends to containerized microservices with auto-scaling
Migrate inventory databases to managed cloud databases with read replicas for real-time sync
Implement API gateway to decouple channel front-ends from backend services
Use cloud-native WAF and tokenization to reduce PCI scope

Outcomes Achieved

60% infrastructure cost reduction through elastic scaling (no over-provisioned peak capacity)
Black Friday handling capacity increased 15x without manual intervention
PCI-DSS audit scope reduced by 70% through tokenization and managed services

Implementation Challenges & Solutions

Challenge: Legacy Dependencies & Proprietary APIs

Issue: Hardcoded system calls to proprietary operating system APIs (e.g., OpenVMS system services, HP-UX kernel extensions) that have no direct x86 equivalent Solution: Implement compatibility shim layers and API abstraction wrappers during migration; for complex cases, engage code refactoring partners to rewrite affected modules Tools: Charon emulation (for VAX/Alpha/SPARC), Micro Focus Enterprise Server (for CICS/JCL), LzLabs SDM Risk Mitigation: Identify API dependencies during Phase 1 discovery; test shim layers thoroughly in PoC

Challenge: Data Format Conversion

Issue: EBCDIC encoding, packed decimal fields, binary-format VSAM datasets, and proprietary date formats that must be converted to ASCII/UTF-8 and relational structures Solution: Build automated ETL conversion pipelines with field-level validation; run parallel output comparison between source and target for a minimum of two full business cycles Tools: Precisely (formerly Syncsort), Ispirer SQLWays, custom Python/PySpark conversion scripts Approach: Convert and validate data in staging environments before cutover; maintain rollback-capable data sync until hypercare completes

Challenge: Performance Parity

Issue: Legacy platforms (especially mainframes) deliver deterministic, low-latency I/O that can be difficult to match with commodity x86 and cloud storage Solution: Conduct detailed MIPS-to-vCPU benchmarking during PoC; leverage NVMe-backed storage, memory-optimized instances, and horizontal scaling to exceed legacy performance Result: In 80%+ of engagements, migrated workloads achieve equal or better throughput at 40-60% lower cost

Challenge: Batch Processing Migration

Issue: Complex JCL job streams with hundreds of steps, conditional logic, and scheduler dependencies that do not translate directly to cloud-native orchestration Solution: Decompose JCL into discrete, containerized job steps orchestrated by modern schedulers; re-engineer serial batch chains into parallel pipelines where data dependencies allow Tools: Apache Airflow, AWS Step Functions, Azure Data Factory, Control-M (cloud edition), BMC Helix

Challenge: Organizational Resistance

Issue: Operations teams with deep legacy expertise may perceive migration as a threat to their roles; business stakeholders fear disruption Solution: Invest in retraining programs that upskill legacy engineers to cloud and DevOps practices; involve operations teams in migration execution so they become the target platform experts Approach: Communicate migration as career development, not displacement; celebrate early wins publicly

Challenge: Regulatory & Compliance Constraints

Issue: Regulated industries require evidence that migrated systems meet the same control objectives as legacy platforms Solution: Map existing compliance controls to cloud-equivalent controls (e.g., CIS Benchmarks, cloud provider compliance programs); engage auditors early with target architecture documentation Frameworks: NIST 800-53, ISO 27001, SOC 2, PCI-DSS cloud supplement, HIPAA cloud guidance

Success Metrics & KPIs

Business Value Metrics

Cost Reduction:
- Infrastructure TCO: 40-70% reduction within 12 months
- Licence costs: 30-60% reduction through open-source adoption
- Support contracts: 50-80% reduction post-decommission
- Energy and facilities: 20-40% reduction (datacenter footprint)

Agility Improvement:
- Provisioning time: From weeks to minutes (95%+ reduction)
- Deployment frequency: 5-20x increase
- Time-to-market for new features: 30-50% faster
- Scaling response: From hours to seconds (auto-scaling)

Risk Reduction:
- Unplanned downtime: 60-80% reduction
- Security patch deployment: From months to days
- DR recovery time: From hours to minutes (cloud-native DR)
- Vendor concentration risk: Eliminated or significantly reduced

Migration Execution Metrics

Delivery Quality:
- Migration success rate: 95%+ (first-attempt cutover)
- Data integrity: 100% validation pass rate
- Performance parity: Within 10% of baseline or better
- Rollback invocations: &lt;5% of waves

Timeline Adherence:
- Wave completion: Within 2 weeks of planned date
- PoC duration: &lt;6 weeks from kickoff to gate
- Total program: Within 15% of original estimate
- Hypercare incidents: &lt;3 P1/P2 per wave

Post-Migration Operational Metrics

Platform Health:
- Availability: 99.9%+ (up from typical 99.5% on legacy)
- Mean time to recovery: &lt;15 minutes
- Automated patching coverage: >95%
- Infrastructure-as-code coverage: >90%

Team Capability:
- Legacy skill dependency: Reduced by 70-90%
- Cloud certification attainment: 80%+ of operations team
- Self-service provisioning adoption: >85%
- Incident resolution time: 40-60% faster

Investment & ROI

Assessment Phase

Model: Fixed-price engagement
Duration: 2-4 weeks
Team: 1-2 senior architects, 1 migration specialist
Deliverables: Application portfolio analysis, migration strategy, target architecture blueprint, business case with 3-year TCO projection
Typical Investment: Scaled to portfolio size; contact us for a scoping conversation

Migration Execution Phase

Model: Time and materials with milestone-based billing, or fixed-price per wave
Duration: 3-9 months depending on portfolio size and complexity
Team: Migration lead, 2-4 migration engineers per wave, data specialist, test engineer
Deliverables: Migrated and validated applications, updated runbooks, decommission plan
Success-Based Component: A portion of fees tied to confirmed cost savings in the first 12 months

Optimization Phase

Model: Monthly retainer or outcome-based pricing
Duration: 3-6 months post-migration
Team: Cloud architect, SRE engineer
Deliverables: Right-sizing recommendations, auto-scaling policies, cost optimization report, modernization roadmap
Typical Savings: Additional 15-25% cost reduction beyond initial migration gains

ROI Timeline

Typical Return Profile:
- Month 1-3: Assessment and PoC investment (net cost)
- Month 4-9: Migration execution; legacy and cloud run in parallel (cost peak)
- Month 10-12: Legacy decommission begins; savings materialize
- Month 13-18: Full run-rate savings achieved (40-70% cost reduction)
- Month 19-36: Cumulative ROI reaches 250-400%

Break-Even:
- Small portfolios (&lt;50 applications): 6-9 months
- Medium portfolios (50-200 applications): 9-14 months
- Large portfolios (200+ applications): 12-18 months

Case Studies

Financial Services Leader

Challenge: Mainframe batch processing system handling 2M+ daily transactions with an 8-hour nightly batch window constraining business operations Solution: Re-platformed core batch workloads to Linux containers on AWS, decomposed monolithic JCL into Airflow-orchestrated microservices Result: 60% cost reduction, batch window reduced to 47 minutes, 24x7 availability with zero unplanned downtime in first 12 months

Global Manufacturer

Challenge: 300+ Unix servers (AIX and HP-UX) across 14 plant sites, each managed independently with inconsistent patching and monitoring Solution: Consolidated to a private cloud on VMware vSphere with centralized management, automated provisioning, and standardized RHEL builds Result: 75% server reduction, unified monitoring and patching, annual savings of $4.2M in support contracts

Healthcare Provider

Challenge: Legacy EMR on proprietary minicomputer platform, unable to support mobile clinician access or FHIR-based interoperability Solution: Migrated EMR to x86 with an API abstraction layer, deployed on Azure with HIPAA-compliant landing zone Result: Enabled mobile access for 2,000+ clinicians, established FHIR APIs for 15 partner integrations, 50% infrastructure cost reduction

Why arqitekta

Deep Platform Experience

We have led migrations spanning every major legacy platform — mainframe, midrange, proprietary Unix, and aging x86 — to modern cloud and virtualized targets. Our architects carry hands-on experience with the specific technical challenges each source platform presents.

Vendor-Neutral Guidance

We are not aligned with any single cloud provider or tooling vendor. Our recommendations are driven by your workload characteristics, compliance requirements, and commercial objectives — not by reseller margins.

Migration Factory Discipline

Our structured, wave-based methodology with standardized runbooks, automation, and quality gates delivers repeatable results. This factory approach reduces per-application migration cost by 30-40% compared to ad-hoc project-based migrations.

Business Continuity Focus

Every migration plan includes validated rollback procedures, parallel-run validation, and hypercare support. We measure success not just by applications moved, but by business disruption avoided.

Re-platforming to Cloud