Section 6 · Data and Analytics Management · 13% of exam
6.3 External data, backup and restore, and import methods
Exam objective
Understand methods to connect to, access, back up, restore, or archive external data; select import tools and external IDs for scenarios.
Choose where data should live based on volume, latency, ownership, and user needs. Salesforce objects store operational records that users and automation work with directly. External objects represent data stored outside Salesforce and can be accessed through Salesforce Connect adapters without copying every record into the org. This can reduce duplication and keep the source system authoritative, but it depends on external availability, adapter support, query constraints, and network behavior. Users still need the correct access to the external data and connected system. Big Objects are designed for very large volumes of data with specific indexing and query patterns. They are not a general replacement for standard objects; transaction, reporting, relationship, and user-interface capabilities differ. Data lakes and warehouses can support large-scale analytics and combine data across systems, while third-party backup or archiving products may provide recovery, retention, and compliance functions. Evaluate what is backed up, how often, how granularly it can be restored, and whether metadata and relationships are preserved.
A backup plan should cover data and metadata, recovery objectives, retention, access to backup copies, and restore testing. A successful backup job does not prove that a usable restore is possible. Test recovery of representative records and relationships in a safe environment. Consider accidental deletion, malicious changes, failed deployments, integration corruption, and org-level incidents. Native export tools may support extracts, but they are not necessarily a complete managed backup and recovery service.
Data Import Wizard is designed for guided imports on supported objects and moderate tasks, with mapping and duplicate-related options. Data Loader supports larger or more technical insert, update, upsert, delete, and export operations, depending on configuration and permissions. Use external IDs when matching incoming records to existing records or related records. An upsert inserts records that do not match and updates those that do, reducing duplicates when the external ID is stable and unique.
Before a bulk import, map source columns to Salesforce fields, normalize values, confirm record ownership and parent lookup strategy, and validate required fields and picklist values. Run a small sample first. Save success and error outputs securely, check counts, and reconcile failures. For parent-child relationships, load parents first or use external IDs in relationship mappings where supported. A partial success can leave the org inconsistent if the follow-up import is not planned.
Data migration decisions should include API limits, batch size, locking, automation volume, duplicate rules, validation rules, and downstream integrations. Large jobs may require a maintenance window or coordination with integration owners. Do not disable validation or automation casually; if a controlled bypass is approved, document scope and re-enable it. Preserve a source extract and rollback plan according to policy. Confirm the user's permissions and data access before choosing an import tool.
Exam focus: use the Import Wizard for guided, supported imports; Data Loader for bulk or more controlled operations; and external IDs for matching and upsert. Use Salesforce Connect when users need external data without copying it, subject to adapter and query constraints. Distinguish operational Salesforce records from Big Objects and warehouses. For backup questions, consider restore testing and the system's recovery requirements rather than assuming periodic export is enough.
Bulk updates can trigger automation, send notifications, or alter access. Test volume and side effects before a production load, and protect exported files that contain personal data.
Recovery planning should define a recovery point objective and recovery time objective with the business. Decide how much data loss is acceptable and how quickly core processes must return. Test restoration of relationships, ownership, files, and metadata—not only a few record values. For imports, prepare a control total, expected error rate, and reconciliation report. Avoid keeping export files on unmanaged devices. If external data is unavailable, decide whether users should see a graceful error, cached value, or an alternate workflow rather than a misleading blank field. Practical scenario
An external billing system remains the source of truth for millions of invoice lines, but support agents need to view selected invoice details alongside Salesforce accounts. Evaluate Salesforce Connect and external objects to avoid duplicating all rows, then verify adapter support, query performance, authentication, and permissions. If agents need to edit and automate the invoice lines as Salesforce records, an external view may not satisfy the requirement and a governed replication design may be needed.
Extra concept
External IDs should be stable identifiers, not display labels that users can change. Define uniqueness, null behavior, and ownership of the key in the source system. A reliable integration contract prevents many duplicate and reconciliation problems.
Key takeaways
- Use external objects when data should remain in its source, subject to adapter constraints.
- Choose import tools by volume and operation; use external IDs for matching.
- A backup plan is incomplete until restore and reconciliation are tested.
- Official references: https://help.salesforce.com/s/articleView?id=005298969&language=en_US&type=1 https://help.salesforce.com/s/articleView?id=sf.data_import_wizard.htm&type=5 https://help.salesforce.com/s/articleView?id=sf.connect_about.htm&type=5 https://help.salesforce.com/s/articleView?id=sf.big_objects.htm&type=5
