The scenario describes a synchronous, real-time interaction where a support agent requires an immediate response to complete a task ("on the spot"). The most appropriate error handling mechanism is for the client system making the API call to handle connection issues or error responses directly from the web service. This "Request and Reply" pattern ensures that if the external verification service is down or returns an error, the failure is caught immediately and a precise error message can be displayed to the agent. This allows the agent to inform the customer and take appropriate action without undue delay.

1. Salesforce Architects

Integration Patterns and Practices Guide. This guide details the Remote Process Invocation—Request and Reply pattern. It states

"The calling process blocks and waits for the response from the remote process." The guide's section on error handling for this pattern emphasizes that the calling application is responsible for handling faults and exceptions returned by the remote service

which directly supports option B. (This is a foundational Salesforce architect document).

2. Salesforce Architects

Architect's Guide to Asynchronous Processing (Trailhead Module). This guide differentiates synchronous and asynchronous processing. The "on the spot" requirement clearly aligns with a synchronous transaction where the user waits for a response

making asynchronous patterns like "fire and forget" (Option D) inappropriate for the primary interaction.

3. Hohpe

G.

& Woolf

B. (2003). Enterprise Integration Patterns: Designing

Building

and Deploying Messaging Solutions. Addison-Wesley. In Chapter 6

this foundational text discusses the "Request-Reply" pattern

noting that the requester blocks until it receives the reply message. It also covers exception handling

where faults from the replier must be managed by the requester

aligning perfectly with the logic in option B.