Salesforce Decode
Salesforcedecode
Back to questions
ApexArchitect (Hardest)ldvperformancearchivingindexing

Architect large data volume strategies for 20M+ record objects

Real World Scenario

Custom Transaction__c object exceeds 18M records. Reports timeout and selective queries fail without custom indexes.

Expected Answer

• Identify selective filter fields; work with Salesforce support for custom indexes • Archive cold data to Big Objects or external warehouse with pointer records • Partition logic by date or region in query filters always leading with indexed fields • Avoid skinny table assumptions—verify query plans in query planner tool • Batch processing ordered by indexed field for efficient chunking • Consider skinny custom indexes on foreign keys used in every query • Capacity planning for storage and API limits with growth projections

Follow-Up Questions & Answers

Click to expand — each follow-up includes a direct, interview-ready answer

Stay within limits by: Identify selective filter fields; work with Salesforce support for custom indexes. Bulkify everything — never query or DML in loops. Partition logic by date or region in query filters always leading with indexed fields. LDV is architectural—retroactive indexing requests take weeks. Document the decision in an ADR and align with enterprise standards.

Architect Perspective

LDV is architectural—retroactive indexing requests take weeks. Plan selective predicates at object design time.