How do you assess the risks and benefits of mainframe migration?

Assessing mainframe migration risks involves meticulous evaluation of existing complexities and dependencies. Identifying potential disruptions, data integrity issues, and ensuring a seamless transition are critical. Benefits include enhanced scalability, cost optimization, and increased agility. Balancing long-term advantages against short-term challenges requires a strategic, phased approach to minimize business impact, maximizing potential for innovation and efficiency gains.

Unearthing complexities and risks: Most enterprises depend on a discovery and assessment phase to unearth risks and assess complexities associated with the migration. While tools can be of great value in this endeavor, there will be blind spots and obscure elements. As decades have gone by, lack of or out of sync documentation, absence of SMEs mean discovery exercises can be tough.

Before starting any migration plan you really need to understand what are the necessary goals that needs to be addressed and what are the steps and tools that are necessary to achieve the goals. Also if you are planning to just migrate everything from your mainframe, stop and think twice. The best approach is to first integrate your mainframe with cloud and then after that check which are the next steps to achieve the plan. Be aware that during the migration plan many surprises will pop up and to try to avoid many of these surprises, many assessments with the migration service provider should be done.

The key objectives to consider when embarking on a mainframe migration project: 1. Reduce Costs: Migrate to lower-cost infrastructure like cloud or x86 platforms to reduce capital and operating expenses 2. Increase Agility: Adopt modern architectures and technologies to enable faster delivery of new capabilities 3. Innovation: Access latest tools and capabilities like AI/ML, containers, etc. to drive innovation 4. Improve Performance: Leverage scalable cloud infrastructure to handle fluctuating workloads 5. Enhanced Security: Implement modern security practices and technologies 6. Sustainability: Reduce energy consumption by migrating to efficient cloud data centres 7. Access Skills: Attract new talent by moving to mainstream technologies

Before starting your migration plan make sure to have a solid understanding of how cloud architecture works and how it’s designed to operate. Make sure you understand about Capex and Opex and have a solid plan based on this. Make sure to have many assessments along with the migration service provider to try to mitigate future surprises that will surly come up thru all the process. Have a real understanding that a mainframe migration plan can become even expansive than modernization in the platform and it can take 1 to 3 or more years. These are some basic points that must be checked and revised before starting any migration plan.

Also, factor in the growth of transactions expected based on business projections . Coupled with organizational policies for data retention and archival, will provide inputs to dynamic and static storage requirements

One thing many miss is the fact that not all data needs to migrate immediately. Partial migrations occur all the time, and there are tools that allow continued access to those difficult programs you cannot migrate in time (in this example it can be any legacy code on any system, but especially on a mainframe such as a zOS box). So when planning, think in terms of partial, staged or full migration if you must go down the path of migrating from one legacy system to another system. Lastly, ask yourself why do I need to migrate off the mainframe to begin with? Mainframes are very useful machines if you utilize accessing them properly.

Before migrating to mainframe, assess if cloud systems can handle your workloads. Calculate data volume for migration, check mainframe app compatibility, and address potential x86 server issues. If possible, consider a modernization approach.

The key considerations are: 1. Application Analysis: Conduct a thorough analysis of existing mainframe applications. 2. Workload Evaluation: Assess mainframe workloads - performance, capacity needs etc. to determine migration approach 3. Migration Strategy: Evaluate different strategies like re-hosting, refactoring etc. based on goals, risk appetite 4. POC: Build a limited POC to demonstrate technical feasibility of plans 4. Target Platform: Assess suitability, compatibility and readiness of target platform like cloud or x86 5. Tools and Resources: Determine availability of suitable tools, skills, partners required for migration 6. Testing and Risks: Develop extensive test plans and evaluate potential risks like failures, outages etc.

I would advise deciding on what the metrics for success look like will be. Use a tool like XINFO to perform a light touch analysis of your system to perform an analysis of your environment:- - Capture CPU time - Storage - Run Times - Other - Look at a snapshot view and use the XINFO tool to perform an Assessment of a before / after view of your environment.

For this it’s my experience it’s not an easy job to get a good view on TCO. Sure we know what amount is at the bottom of the hardware, software and support contracts. Sure we know how many fte’s are working on for the platform. But usually when you take a good look at what’s “in the books” tagged as “for the mainframe” (or budgets) you’ll find some interesting facts. As a side note : usually TCO is not the real driving factor for migrations yet often used as a “fake argument” Depending on your workloads and size it’s my honest experience it runs cheaper on the Mainframe (if you look at “real” cost). When you add the NFRs like carbon footprint, energy consumption and transactions per square feet of datacenter, etc it’s easy math

Think about your budget before moving from mainframe to the cloud, that would be expensive. If you're going all-in on the cloud, make sure it's worth the investment.

The key points are: 1. Current Mainframe Costs: Thoroughly analyze current costs-hardware, software, licensing, operations, maintenance etc. 2. Migration Costs: Estimate all one-time and recurring costs of migration effort-tools, services, testing, training etc. 3. Target Platform Costs: Evaluate ongoing infrastructure and support costs on new platform post migration 4. Cost Savings: Compare current and future costs to determine potential savings from migration 5. Additional Benefits: Consider indirect benefits like agility, innovation, sustainability delivered 6. Risks and Uncertainties: Factor in potential risks like failures and delays that can increase costs 7. Timeline: Analyze short-term costs vs long-term TCO and ROI over 3-5 years.

The key considerations are: 1. Skills Assessment: Evaluate skills of teams and identify gaps for training/hiring 2. Executive Buy-in: Ensure leadership support to facilitate change management 3. Staff Engagement: Communicate plans and address concerns of impacted staff 4. Process Changes: Assess process changes needed to support new platforms/ways of working 5. Partners and Vendors: Determine partner and vendor readiness to support migration 6. Financial Planning: Develop budgets and plans to fund migration costs 7. Risk Management: Identify risks and develop mitigation plans

Organization readiness: Organization readiness also includes sustained availability of key SMEs to support complex migration activities without impacting business as usual. In many scenarios establishing a baseline from where migration starts is a non-trivial activity and needs significant efforts from all stakeholders. While adequate buffer for cost and timeliness would have been provisioned during estimation, resilience to handle overruns and agility to adopt to the unforeseen is also critical.