1. Scenario: Multi-Domain Expansion

Q: Your company currently runs Product MDM in STEP. Management wants to add Supplier and Location domains. What architectural risks do you highlight?
A:

• Cross-domain dependencies may introduce circular validation rules if Product → Supplier → Product loops occur.

• Existing workflows may not be modular; adding new domains may require refactoring them into domain-specific processes.

• Performance risk: A shared repository may overload the JVM heap if all domains use heavy inheritance models.

• Harmonization rules may conflict because entity structures differ across domains.

• Governance bodies may not be ready to manage multi-domain stewardship, causing delays in approvals.

2. Scenario: Multi-Hierarchy Explosion

Q: A retail client wants the same product to appear in Sales, Web, and Internal hierarchies. How do you avoid hierarchy explosion?
A:

• Use Classification hierarchies for marketing-driven taxonomies.

• Use alternative structures instead of duplicating base records.

• Store hierarchy-specific attributes in hierarchy assignment nodes.

• Leverage dynamically built “virtual hierarchies” using rules instead of manually maintaining static ones.

3. Scenario: ERP Integration Fails Randomly

Q: STEP → SAP integration intermittently fails due to inconsistent payload sizes. Root cause?
A:

• SAP limits message size; large STEP records with collections exceed it.

• STEP's Data Load Service may send uncompressed payloads.

• Multivalue attributes may create oversized XML nodes.

• Solution: Implement outbound throttling, compress large payloads, and apply pagination.

4. Scenario: Conflict Between Workflow & Business Rule

Q: Workflow approves a record but a validation rule blocks publishing. What do you fix?
A:

• Check rule sequencing; validation rules may be evaluated after workflow completion.

• Add workflow tasks to enforce rule compliance earlier.

• Create rule exceptions for workflow-approved data.

• Or enforce a “gatekeeper rule” pre-approval.

5. Scenario: STEP Cache Not Updating

Q: You update attribute values via a batch process, but UI still shows old data. Why?
A:

• STEP Cache or Full-Text Search index not refreshed.

• Verify incremental indexing jobs.

• Cache invalidation rules may be misconfigured.

• Restart Metadata service or apply forced reindex.

6. Scenario: Crosswalk Deduplication Failure

Q: Despite same GTIN crosswalk, duplicates still occur. Why?
A:

• GTIN crosswalk may not be configured as a unique key.

• Multiple GTIN formats exist (with/without leading zeros).

• Matching rules may be disabled for certain load channels.

• Solution: Normalize GTIN formats, enforce uniqueness at ingestion, and apply survivor rules.

7. Scenario: Massive Import Failing Midway

Q: 1M records imported; after 60%, job fails. Logs show “heap space.” What do you optimize?
A:

• Increase JVM heap on batch and metadata servers.

• Change import to chunk mode.

• Remove expensive rule evaluations during import.

• Disable full indexing until import completes.

8. Scenario: Slow Workflow Approvals

Q: Large workflow approval steps take minutes. What do you optimize?
A:

• Review large collections in tasks; reduce payload size.

• Clean old workflow instances.

• Scale workflow engine on separate nodes.

• Simplify gateways and conditions.

9. Scenario: Orphaned Classification Nodes

Q: Duplicate import created classification nodes without parents. How do you resolve?
A:

• Use hierarchy validation report to find orphans.

• Auto-merge nodes based on code or name similarity.

• Lock hierarchy before fixing to prevent new orphans.

10. Scenario: Product Variant Explosion

Q: Brand creates 200 variant SKUs per base product. How do you prevent performance issues?
A:

• Use Product Configurator instead of real SKU entities.

• Use inheritance to avoid repeated attributes.

• Avoid loading all variants into workflow tasks.

11. Scenario: Reference Model Conflict

Q: Supplier and Product entities reference each other. How to avoid cyclic dependencies?
A:

• Split relationships: Product → Supplier, Supplier → ProductCategory.

• Use lookup tables.

• Avoid bidirectional hard references.

12. Scenario: Migrating from STEP 9 to 11

Q: What key risk areas to validate?
A:

• Deprecated API endpoints.

• Changed workflow engine behavior.

• New caches.

• Rule engine performance shifts.

13. Scenario: Attribute Level Security

Q: Only finance should see margin attributes. What’s the best design?
A:

• Use STEP access control lists on attributes.

• Create finance-specific workflows.

• Restrict full-text search indexing for sensitive fields.

14. Scenario: Dangerous Match Rule

Q: Match rule incorrectly merges unrelated customer records. How do you fix?
A:

• Reduce similarity weights.

• Add mandatory match keys.

• Create manual exceptions workflow.

15. Scenario: Supplier Onboarding Automation

Q: Supplier must submit docs; STEP must validate them. How?
A:

• Use inbound API for doc upload.

• Validate metadata with rules.

• Route to approval workflow.

16. Scenario: Attribute Change Triggers Wrong Publish

Q: A minor attribute triggers full record republish. Why?
A:

• Attribute wrongly included in the publish trigger set.

• Change tracking settings too broad.

• Fix delta publish logic.

17. Scenario: Golden Record Derivation Failing

Q: Survivor rules not applied consistently. Root causes?
A:

• Rule order incorrect.

• Channel priority misconfigured.

• Crosswalk entered after rule execution.

18. Scenario: API Response Too Slow

Q: STEP API takes 15 seconds for simple record fetch. What do you check?
A:

• Check custom filters causing slow DB scans.

• Reduce number of relations returned.

• Add caching layer.

• Increase DB connection pool.

19. Scenario: External System Requires Real-Time Sync

Q: Partner system expects near real-time updates. How do you design?
A:

• Use STEP Event Management.

• Queue events in Kafka.

• Build microservices on top of event streams.

20. Scenario: Wrong Data Pushed to Downstream

Q: Syndicator pushes partially enriched product data. What do you inspect?
A:

• Wrong workflow state mapped to publish.

• Syndication rule missing required attributes.

• Failed job not raising errors.

21. Scenario: Memory Leak in Custom Java Plugin

Q: JVM memory rises until crash. Why?
A:

• Streams not closed.

• Large collections held statically.

• Excessive logging.

22. Scenario: SQL Backend Lock Contention

Q: Many STEP transactions blocked. Root cause?
A:

• Long-running queries from reporting tools.

• Bulk updates without batching.

• Missing DB indexes.

23. Scenario: Reference Data Not Versioned

Q: CoA values changed, breaking old workflows. How to fix?
A:

• Use version-controlled reference entities.

• Freeze old versions for past records.

24. Scenario: Metadata Deploy Fails

Q: Deployment errors due to “unresolvable dependency.” Why?
A:

• Invalid entity references.

• Deleted attribute still referenced in workflow.

• Cyclic inheritance.

25. Scenario: Product Lifecycle Misalignment

Q: Product marked “Discontinued” but appears on website. Why?
A:

• Wrong syndication rule.

• Website reading from staging, not published.

• Lifecycle state not mapped to web availability.

26. Scenario: ETL Bulk Load Creates Dirty Data

Q: Imports bypass rules. How do you enforce validations?
A:

• Enable RuleCheckDuringImport.

• Put preprocessing pipeline.

• Post-load cleanup workflows.

27. Scenario: Rule Engine Bottleneck

Q: CPU spikes due to heavy rules. Optimization?
A:

• Replace rule scripts with derivation attributes.

• Use downstream calculation instead.

• Limit rule scope.

28. Scenario: Multi-Language Inconsistency

Q: Product names missing in French. How to handle?
A:

• Use completeness rules.

• Auto translate with external API.

• Create language-specific steward queues.

29. Scenario: Media Asset Explosion

Q: Tens of thousands of images overloaded asset server. Solution?
A:

• Use external DAM.

• Maintain only metadata in STEP.

• Store URLs not binaries.

30. Scenario: Circular Approval Loops

Q: Workflow returns to same user repeatedly. Cause?
A:

• Broken gateway conditions.

• Hidden mandatory fields.

31. Scenario: Multi-Region Data Sovereignty

Q: EU and US require different governance. How to design?
A:

• Regional data domains.

• Regional workflows.

• Localized access control.

32. Scenario: Competing Survivorship Needs

Q: Sales wants highest priority; Finance wants strict ERP override. Solution?
A:

• Layered survivorship (Finance → ERP → Sales).

• Channel-specific rules.

• Override workflow.

33. Scenario: Classification Migration

Q: Moving from UNSPSC to custom taxonomy. Risks?
A:

• Massive reclassification effort.

• Loss of historical reporting.

• Mapping inconsistencies.

34. Scenario: Multiple ERP Connectors

Q: Product must be sent to SAP & Oracle with different schemas. How?
A:

• Multiple syndication targets.

• Attribute mapping layers.

• Delta publish logic.

35. Scenario: Data Stewards Overloaded

Q: Too many manual tasks. How do you reduce workload?
A:

• Pre-validate with automated rules.

• Auto-assign categories.

• Introduce guided workflow tasks.

36. Scenario: Unstable Imports

Q: Daily feed inconsistent; some rows missing. Solution?
A:

• File completeness check.

• Hash validation.

• Retry mechanism.

37. Scenario: Multi-Country Unit Conversions

Q: US uses ounces; EU uses grams. Solution?
A:

• Standard base unit storage.

• Convert dynamically via rules.

• Country-specific views.

38. Scenario: Record Locking Issues

Q: Users get “record locked” frequently. Why?
A:

• Abandoned sessions.

• Long workflow assignments.

• High concurrency editing same entity.

39. Scenario: Inheritance Broken

Q: Child product not inheriting new attributes. Why?
A:

• Attribute not included in parent template.

• Broken inheritance link.

• Disabled “propagate changes”.

40. Scenario: Workflow Restart Needed

Q: Big product changes require workflow rerun. How?
A:

• Add workflow restart trigger.

• Use validation to auto-route.

41. Scenario: Multi-Channel Publish Rules

Q: Web channel needs rich content; ERP needs minimal. Solution?
A:

• Separate publish templates.

• Channel-specific completeness rules.

42. Scenario: Media URLs Broken

Q: CDN migration broke all media links. How to fix?
A:

• Bulk URL update via script.

• Sync check to ensure new metadata valid.

43. Scenario: Attribute Life-Cycle

Q: An attribute is relevant only in Packaging stage. How to manage?
A:

• Lifecycle-based attribute visibility.

• Conditional UI configurations.

44. Scenario: New Business Rule Slows System

Q: Rule evaluation adds 800ms per save. Fix?
A:

• Convert to pre-cached lookup.

• Reduce relational traversals.

• Batch validations.

45. Scenario: Audit Log Too Large

Q: Audit log approaching TB size. Actions?
A:

• Archive old logs.

• Compress.

• Move to cold storage.

46. Scenario: Crosswalk Conflicts

Q: Two systems use same external ID range. Prevent conflict?
A:

• Add namespace prefix.

• Separate crosswalk types.

• Enforce naming convention.

47. Scenario: Multi-Threaded Import Corruption

Q: Parallel import corrupts parent-child structure. Why?
A:

• Race conditions.

• Missing import dependencies.

• Solve using staged imports.

48. Scenario: Workflow Routing Based on Attribute

Q: Workflow should route based on Brand. Error occurs when Brand changed. Why?
A:

• Workflow context cached old value.

• Must refresh context.

• Add dynamic decision gateway.

49. Scenario: Step Distribution Server Overloaded

Q: Large bulk exports overload server. How to optimize?
A:

• Offload export to separate node.

• Use streaming export.

• Increase JVM buffer.

50. Scenario: Disaster Recovery

Q: What is your DR plan for STEP?
A:

• Active-passive DB replication.

• Backup metadata snapshots.

• Automated redeploy scripts.

• DR test every quarter.