Cloud service · AWS · Updated

AWS Mainframe Modernization - AWS Mainframe Modernization provides tools and runtime options for assessing, migrating, refactoring, and operating mainframe applications on AWS.

In short

The service supports automated refactoring into modern application stacks and replatforming with compatible mainframe runtimes, together with development, data transfer, and application testing capabilities. Its managed runtime experience is closed to new customers, and new customer access to the self-managed experience closes effective June 30, 2026; existing customers can continue using their respective experiences.

Cloud Services Hub →

9 key features of AWS Mainframe Modernization

01
Application assessment and dependency analysis

Assessment capabilities help identify application boundaries, dependencies, and modernization scope. Use the results to identify missing artifacts, group related programs, and plan migration waves before choosing a runtime.

02
Automated application refactoring

AWS Transform for mainframe capabilities support analysis and conversion of supported mainframe workloads into modern application stacks. Validate generated applications against the original business behavior and review unsupported constructs and external dependencies.

03
Mainframe application replatforming

Rocket Software, formerly Micro Focus, technology supports recompiling and running compatible mainframe applications with limited source changes. Confirm support for the application's COBOL or PL/I constructs, transaction processing, JCL, and data access requirements.

04
Managed runtime operations

Existing managed runtime customers can deploy applications into monitored runtime environments with high-availability options and managed infrastructure operations. Application definitions, runtime engines, capacity, networking, and dependent storage still require workload-specific configuration.

05
Mainframe data set transfer

File Transfer with BMC supports data set discovery, selection, and transfer to AWS. Choose code page conversion for supported text-processing workflows or binary transfer when tools must process the original mainframe representation.

06
Functional equivalence testing

Application Testing automates supported functional equivalence and regression testing workflows. Use representative inputs and expected outputs to detect behavioral differences before cutover and after subsequent application changes.

07
Batch execution and recovery

Managed application operations include submitting and canceling batch jobs. Supported refactored applications can restart a job at a specified step, allowing more targeted recovery after operators verify restart safety.

08
Data set export

Managed runtimes for refactored and replatformed applications can export data sets to Amazon S3, with optional encryption. Use exports to move data between environments or retain archival copies under appropriate access controls.

09
Assembler-to-COBOL conversion

Code Conversion with mLogica provides automated conversion of supported Assembler code to COBOL. This can remove a language dependency that otherwise blocks refactoring or replatforming, but the converted code still requires compilation and functional validation.

Where AWS Mainframe Modernization is used

  1. 01
    Replatform an existing COBOL workload

    Move a compatible transaction and batch application to a Rocket Software runtime while retaining much of its source code. Validate JCL behavior, file semantics, database access, and operational procedures before replacing the mainframe execution path.

  2. 02
    Refactor applications into Java

    Convert supported mainframe applications into a modern Java-based stack using automated transformation capabilities. Establish functional parity first, then make incremental architectural changes rather than combining migration with an unrestricted redesign.

  3. 03
    Validate modernization releases

    Use Application Testing to compare supported migrated application behavior with a mainframe baseline. Repeat representative transaction and batch scenarios during migration and after changes to code, runtime versions, or data mappings.

  4. 04
    Prepare mainframe data for migration

    Discover and transfer mainframe data sets to AWS for conversion, testing, or loading into target systems. Select binary transfer when exact source bytes must be retained, and verify encodings and record layouts before conversion.

  5. 05
    Operate migrated batch processing

    For eligible existing runtime customers, execute migrated batch jobs and integrate them into operational scheduling and monitoring. Define recovery procedures that account for step dependencies, committed updates, and duplicate output.

  6. 06
    Remove Assembler migration blockers

    Convert supported Assembler components to COBOL with mLogica before a broader modernization effort. Prioritize components on critical application paths and test numeric behavior, interfaces, and performance after conversion.

How AWS Mainframe Modernization is priced

Managed runtime consumption

Eligible existing customers pay for provisioned managed runtime capacity according to the selected engine, instance configuration, Region, and applicable rates. Include all instances in high-availability environments and do not assume that an idle deployed environment is free.

Self-managed infrastructure and runtime licensing

Self-managed deployments incur charges for the AWS infrastructure they use and the applicable runtime software pricing or licensing arrangement. AWS Blu Age Runtime non-managed offers pay-as-you-go billing; verify the terms for the selected runtime and deployment model.

Capability-specific charges

Development environments, file transfer, application testing, and code conversion can have separate pricing terms or usage charges. Estimate each selected capability independently rather than assuming it is included in runtime pricing.

Supporting AWS resources

Storage, databases, monitoring, encryption keys, networking, and data transfer are billed under their respective AWS services. Include test environments, retained artifacts, log retention, and migration rehearsals in the total cost estimate.

Services AWS Mainframe Modernization works with

A common architecture separates modernization artifacts, runtime execution, and persistent data. Amazon S3 holds deployment packages, transferred mainframe files, and exported data sets. Eligible managed runtime customers deploy applications into VPC-connected environments, while compatible Amazon RDS databases provide relational persistence where required by the target application. Database selection and schema conversion depend on the runtime engine and migration design.

IAM roles define deployment and runtime access, and AWS KMS provides encryption controls for supported resources and data workflows. Use distinct roles and storage locations for development, testing, and production. Private subnets, restrictive security groups, and supported interface endpoints limit network exposure, but application transaction endpoints require their own connectivity design.

CloudFormation makes supported infrastructure definitions repeatable, while CloudWatch provides logs and monitoring for runtime operations. For example, a release process can provision infrastructure, publish an application package to S3, deploy the application, and run functional regression tests before production promotion. Data loading, batch restart safety, database recovery, and rollback remain explicit engineering responsibilities rather than consequences of infrastructure automation.

Amazon S3Amazon S3

Amazon S3 stores deployment artifacts and supports mainframe data transfer and managed runtime data set export workflows. For example, export a data set to a restricted bucket before loading it into a test environment. Apply encryption, retention policies, and separate permissions for application artifacts and sensitive business data.

Amazon CloudWatchAmazon CloudWatch

Managed runtimes integrate with CloudWatch for operational monitoring and application logs. For example, configure a log metric filter for a known batch failure message and create an alarm for the operations team. Set retention periods explicitly and avoid logging sensitive transaction payloads.

Amazon VPCAmazon VPC

Managed runtime environments use VPC networking to connect applications with clients and supporting services. For example, place a runtime in private subnets and permit database traffic only on the required port. Where supported, use AWS PrivateLink interface endpoints for private service API access; API connectivity and application transaction connectivity are separate paths.

AWS CloudFormationAWS CloudFormation

CloudFormation supports repeatable provisioning of supported modernization resources and their surrounding infrastructure. For example, define a runtime environment, application resource, IAM roles, and security groups in reviewed templates. Keep environment-specific settings parameterized and manage application data migration separately from infrastructure creation.

Amazon RDSAmazon RDS

Modernized applications can use compatible Amazon RDS databases according to their runtime engine and application architecture. For example, a refactored Java application can connect to an RDS for PostgreSQL database after its schema and data access mappings have been migrated. Validate transaction isolation, character encoding, connection limits, and recovery behavior rather than assuming mainframe database semantics carry over unchanged.

What is AWS Mainframe Modernization

2 min read

AWS Mainframe Modernization supports the migration and operation of mainframe workloads on AWS. Its capabilities cover application assessment, code transformation, development, data movement, functional testing, and runtime operations. A modernization project still requires application discovery, dependency mapping, target architecture design, and validation of business behavior; moving source code alone does not migrate an operational workload.

The service supports two principal modernization approaches. Automated refactoring uses AWS Transform for mainframe capabilities to convert legacy applications into modern languages and frameworks, including Java-based applications. Replatforming uses Rocket Software, formerly Micro Focus, technology to run compatible mainframe applications on AWS while retaining much of the existing source code. Select an approach after assessing language support, transaction dependencies, batch processing, data formats, and the amount of application change the team can support.

Access restrictions are important when selecting a target architecture. The managed runtime environment experience is no longer open to new customers. New customer access to the self-managed experience closes effective June 30, 2026. Existing customers can continue using the respective experiences, but AWS does not plan new features for the self-managed experience while continuing security and availability improvements. Confirm eligibility and regional availability before designing a deployment around either experience.

For existing managed runtime customers, an application represents a migrated workload and is associated with a runtime environment. The application specifies an engine appropriate to the modernization approach. Engineers deploy application versions, configure supporting resources, import data sets, submit batch jobs, and monitor execution. High-availability configurations and runtime sizing must reflect transaction throughput, batch windows, recovery objectives, and engine-specific requirements rather than mainframe capacity measurements alone.

Data migration requires explicit treatment of mainframe encodings and record structures. File Transfer with BMC can discover mainframe data sets and transfer selected data to AWS, with code page conversion options or binary transfer when downstream tools require the original representation. Managed runtime data sets can also be exported to Amazon S3, optionally with encryption. Validate record counts, packed decimal fields, character conversion, and application-level totals before accepting a migration.

Application Testing supports automated functional equivalence and regression testing. Build representative test cases for online transactions and batch processing, then compare results against established mainframe behavior. Differences in sorting, numeric precision, date handling, and file encoding deserve targeted checks. For supported refactored applications, restarting a batch job at a specific step can assist recovery, but operators must still assess previously committed database changes and external side effects before rerunning work.

Production architectures normally combine private networking, least-privilege IAM roles, encrypted storage, centralized logging, and repeatable deployment definitions. Separate development, testing, and production resources, and keep deployment artifacts distinct from production data. Budget for runtime capacity, storage, databases, data transfer, optional modernization tools, and testing rather than treating the service as a single migration charge.

AWS architecture icons are the property of Amazon Web Services. BlowStack is not affiliated with or endorsed by AWS.