Salesforce Decode
Salesforcedecode
Back to questions
Data CloudArchitect (Hardest)disaster-recoverybackupmetadata

Architect disaster recovery for Data Cloud metadata and mappings

Real World Scenario

Accidental deletion of data model and segment definitions during sandbox refresh planning causes week-long rebuild in production.

Expected Answer

• Export metadata definitions regularly to git-backed documentation • Separate sandbox refresh policy for Data Cloud config vs CRM • Runbook for rebuilding streams from infrastructure-as-code or checklist • Role-based delete permissions on production config • Test restore procedure in DR drill annually • Version segment definitions with change log • Maintain dependency diagram stream → DMO → segment → activation

Follow-Up Questions & Answers

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

Direct answer: Export metadata definitions regularly to git-backed documentation Also consider: Separate sandbox refresh policy for Data Cloud config vs CRM In practice: Runbook for rebuilding streams from infrastructure-as-code or checklist Document the decision in an ADR and align with enterprise standards.

Architect Perspective

Data Cloud config is code — treat deletes like production schema drops with approvals.