According to the exam guide, one of the objectives is to "describe the use cases and considerations for using various tools and techniques for data migration (for example, Data Loader, Bulk API)"1. This implies that option B is the correct way to extract large volumes of data from Salesforce. The Bulk API is designed to handle large-scale data operations and avoid timeouts. Option A is not correct because the SOAP API is not optimized for large data sets and may encounter limits. Option C is not correct because GZIP compression does not prevent timeouts, but rather reduces the size of the data transferred. Option D is not correct because Batch Apex is used to process records asynchronously in Salesforce, not to export data to an external system.

The issue that an architect will have when implementing the quarterly accounts receivable report in Salesforce is that Salesforce does not allow more than four objects in a single report type. A report type defines the set of records and fields available to a report based on the relationships between a primary object and up to four related objects. A report type has the following limitations:
It cannot include more than four objects in a single report type, which means that the report cannot compile data from five objects (Accounts, Contacts, Opportunities, Orders, and Order Line Items) at once.
It cannot include objects that are more than two relationships away from each other, which means that the report cannot access fields from Order Line Items through Opportunities and Orders.
It cannot include objects that have a many-to-many relationship with each other, which means that the report cannot access fields from Contacts and Opportunities through the junction object Opportunity Contact Role.