DATA-ARCHITECT Practice Questions
Prepare for DATA-ARCHITECT with more than an answer.
- Exam fee
- $400 USD
- Level
- Architect
- Valid for
- No expiration (requires maintenance exams)
Domains covered on the exam 6
- Salesforce Data Management25%
- Data Modeling / Database Design25%
- Large Data Volume Considerations20%
- Data Migration15%
- Data Governance10%
- Master Data Management5%
- 1
During a large data migration of millions of records with complex ownership and sharing rule dependencies, the migration is running extremely slowly. The architect suspects that the constant, synchronous recalculation of sharing rules after each batch is the primary bottleneck.
What is the most effective way to improve performance in this scenario?
Show answer details
Correct answer: B
This is the correct approach. The Deferred Sharing feature is designed for this exact purpose. An administrator can suspend sharing calculations before the migration begins, which prevents the system from re-evaluating access for every record loaded. After the migration is complete, the calculations can be resumed and run as a single, large background process, dramatically speeding up the data load.
- 2
An architect is designing a data model for a recruiting application. There will be an
Application__cobject and aReview__cobject. Multiple reviews must be associated with each application. If an application record is deleted, all of its associated review records must also be deleted automatically.Which type of relationship should be created on the
Review__cobject to link it to theApplication__cobject?Show answer details
Correct answer: B
A master-detail relationship is a tightly coupled relationship where the detail (child) record's existence is dependent on the master (parent). This relationship type provides a cascading delete feature, meaning if the master
Application__crecord is deleted, all detailReview__crecords are automatically deleted as well. This perfectly matches the requirement. - 3
An administrator wants to ensure that the 'Close Reason' text field on a Case is always filled out by a user when the Case Status is changed to 'Closed'.
What is the most appropriate declarative tool to enforce this business rule?
Show answer details
Correct answer: B
A validation rule is the perfect tool for this requirement. It evaluates a formula and, if it returns true, prevents the user from saving the record and displays an error message. The formula would check if the Status is 'Closed' AND the 'Close Reason' field is blank (e.g.,
AND(ISPICKVAL(Status, 'Closed'), ISBLANK(Close_Reason__c))). - 4
A company has accumulated over 100 million Task records in their Salesforce org. These records are essential for historical auditing but are rarely accessed after 24 months. Their data storage is nearing its limit, and the large number of Tasks is degrading the performance of activity-related reports and SOQL queries.
What is the most appropriate data archiving strategy for these historical Task records?
Show answer details
Correct answer: A
This is an excellent on-platform archiving strategy. Big Objects are designed to store large volumes of data for archival and audit purposes at a lower cost and without impacting the performance of primary objects. Moving the data to a Big Object frees up data storage, improves the performance of the primary Task object, and keeps the data accessible within Salesforce for auditing via Async SOQL.
- 5
A financial services company needs to implement a data governance program to comply with new regulations. They must classify all fields containing customer data based on their sensitivity level (e.g., Public, Internal, Confidential, PII). This classification needs to be stored as metadata against each field and be accessible for reporting and compliance audits.
Which standard Salesforce feature should be used to meet this requirement?
Show answer details
Correct answer: C
Salesforce provides standard metadata fields directly on the field definition for Data Classification. These include
Compliance Categorization,Data Owner,Data Sensitivity Level, etc. This feature is designed precisely for this use case, allowing administrators to tag fields with governance metadata that can be reported on and used for compliance. - 6
When implementing an MDM solution, two duplicate Contact records are identified for 'Jonathan Smith'.
- Source A (CRM): Phone = 555-1234, Last Modified = 2 days ago.
- Source B (Billing): Phone = 555-5678, Last Modified = 10 days ago.
The business has defined a rule that the most recently modified record should be the source of truth for the phone number. However, if a phone number is missing from the most recent record, the value from the older record should be used.
What is this type of MDM rule called?
Show answer details
Correct answer: B
This is a classic example of a Data Survivorship Rule (or 'merge rule'). It defines the logic for determining which attribute values from which source records 'survive' to become part of the final, merged golden record. The rule described (most recent, with a fallback) is a common survivorship pattern.
- 7
A global retailer, Cloud Kicks, has over 200 million sales transaction records stored in a custom
Sales_Transaction__cobject, which is a child of theAccountobject. The executive team needs a dashboard that displays year-over-year sales trends, aggregated by product category and region. The underlying reports for this dashboard are consistently timing out due to the massive data volume and the required aggregations.Data Model:
Account(Region, etc.)Sales_Transaction__c(Date, Amount, Product_Category__c, Lookup to Account)
Challenge:
- Aggregating 200M+ records in real-time for dashboard components is not performant.
- The business needs trend analysis, not individual transaction details, on the dashboard.
Which solution should the architect design to provide performant trend reporting for the executive dashboard?
graph TD subgraph Nightly Process A(Sales_Transaction__c) -- SOQL Query --> B{Aggregate Data} B -- Group by Category, Region, Month --> C(Update or Insert) end subgraph Reporting Layer D[Custom_Summary__c] --> E(Executive Dashboard) end C --> DShow answer details
Correct answer: B
This is the correct and most scalable architectural pattern for reporting on large data volumes. Instead of querying millions of raw records at runtime, the dashboard queries a small number of pre-aggregated summary records. This approach is highly performant, avoids timeouts, and provides the exact level of detail needed for trend analysis. The batch job handles the heavy lifting offline.
- 8
A global MedTech company, Vitality Health, is consolidating three regional Salesforce orgs (AMER, EMEA, APAC) into a single global org. Each region has a custom
Patient__cobject, but the data models have critical differences in fields, validation rules, and relationships to a customMedical_Device__cobject. The architect must design a unified, scalable data model that complies with GDPR and HIPAA, tracks device service history efficiently, and supports an anticipated volume of over 80 million patient records within two years.Key Requirements:
- A single, global representation of a Patient.
- Strict compliance with regional data privacy laws (GDPR/HIPAA).
- A scalable many-to-many relationship between Patients and their assigned Medical Devices, including a detailed service history for each device assignment.
- High performance for patient record retrieval and reporting.
Which data model should the architect recommend to meet these complex requirements?
Show answer details
Correct answer: A
This is the optimal solution. Person Accounts provide a standard, scalable way to represent individuals, combining Account and Contact functionality, which is ideal for a Patient model. A junction object (
Device_Assignment__c) correctly models the many-to-many relationship and can hold rich details like service history. Record types on the Person Account effectively manage regional variations in layouts and processes, while standard field encryption and security features can be leveraged for GDPR/HIPAA compliance. - 9
Nimbostratus Solutions has a custom object
Project__cwith over 50 million records, which is the child in a master-detail relationship to the standard Account object. A nightly batch job that updates aStatus__cfield on a large subset ofProject__crecords is frequently failing withUNABLE_TO_LOCK_ROWerrors. Investigation reveals that many of the updatedProject__crecords are children of a few very large Accounts.What is the most effective approach to mitigate these locking errors during the batch update?
Show answer details
Correct answer: B
This is the recommended best practice for mitigating lock contention caused by parent-child data skew. By ordering the child records by their parent ID (
AccountId), the batch job processes all children for a single parent sequentially. This prevents different parallel batches from attempting to lock the same parent Account record simultaneously, thus avoiding theUNABLE_TO_LOCK_ROWerror. - 10
An e-commerce company, ShopSphere, needs to store customer clickstream data and IoT device events in Salesforce. The anticipated data volume is several billion records per year. The data must be queryable for large-scale trend analysis and machine learning models, but it does not need to support standard sharing rules, triggers, or workflows. The solution must be native to the Salesforce platform.
Which TWO Salesforce technologies should an architect recommend for storing and querying this data? (Select TWO)
Show answer details
Correct answer: B, E
Big Objects are specifically designed to store and manage massive data volumes (billions of records) on the Salesforce platform. They are ideal for archival, event monitoring, or historical data that doesn't require the full feature set of standard or custom objects.
Async SOQL is the primary mechanism for running queries against Big Objects. It is designed to handle the massive scale of data stored in Big Objects by running the query in the background and storing the results in a target object.
