📖 What is Parallel Running?
Parallel Running is a system conversion strategy where the legacy system and the new system operate simultaneously for a set period. This allows for the verification of new system outputs against known correct results.
"While this is the safest conversion method, it is also the most expensive because it requires double the effort and resources to maintain two systems."
📚 Certification: Certified Information Systems Auditor (CISA)
🔑 What are the Key Concepts of Parallel Running?
- ▸ Verification and Validation: Comparing outputs from the new system against the legacy system to ensure data accuracy and functional reliability before final cut-over.
- ▸ Risk Mitigation: Minimizing business disruption by maintaining a fully operational fallback system, ensuring continuity if the new system encounters critical failures.
- ▸ Resource Intensity: Increasing operational costs and workload because staff must perform double data entry and maintain two separate environments simultaneously.
- ▸ Cut-over Criteria: Defining specific, measurable benchmarks and a timeframe to determine when the new system is stable enough to decommission the legacy system.
- ▸ Data Synchronization: Ensuring identical input data is processed by both systems to allow for an accurate, side-by-side comparison of the resulting outputs.
🎯 How does Parallel Running appear on the CISA Exam?
You may be asked to identify the most risk-averse conversion strategy for a mission-critical financial system where any output error could lead to significant regulatory penalties or financial loss.
A scenario might describe an organization with a limited budget and staff; you will need to evaluate whether parallel running is feasible given the high cost of double data entry.
Expect questions where you must analyze a failed system implementation and determine if a parallel run would have identified the defects before the legacy system was prematurely shut down.
❓ Frequently Asked Questions
When should a CISA recommend parallel running over a 'big bang' approach?
Recommend parallel running for mission-critical systems where the cost of failure is extreme. Big bang is only appropriate for non-critical systems or when the legacy system is completely incompatible with the new one.
What is the primary audit concern during a parallel run?
The primary concern is data integrity and 'double entry' errors. Auditors check if the same data was entered into both systems and if discrepancies were formally investigated and resolved.
How is the end of a parallel run determined?
The run concludes once the new system consistently produces outputs that match the legacy system's results over a predefined period, meeting all agreed-upon acceptance criteria and stakeholder sign-offs.