You have a transactional application that stem data in an Azure SQL managed instance. When should you implement a read-only database replica?

The primary use case for a read-only database replica is to implement read scale-out. Transactional (OLTP) workloads are often impacted by long-running, resource-intensive analytical or reporting queries. By creating a read-only replica, these read-heavy workloads can be redirected away from the primary, transactional database. This isolation prevents reporting queries from consuming resources (CPU, I/O, memory) and holding locks that would otherwise slow down or block the main application's transactions, thus ensuring stable performance for the transactional workload.

B. Auditing is a built-in feature of Azure SQL Managed Instance that tracks database events and writes them to a log. It does not require a separate read-only replica to function.
C. A replica for a regional outage is primarily for disaster recovery (DR), not just read-scaling. While a geo-secondary replica can be readable, its main purpose is to provide a failover target to ensure business continuity.
D. The Recovery Point Objective (RPO) relates to the maximum acceptable data loss. While replication is involved, implementing a replica is not the direct method to improve RPO; that is typically managed by backup frequency and transaction log shipping intervals.

1. Microsoft Learn

Auto-failover groups - Azure SQL Database & Azure SQL Managed Instance.

Reference: Under the section "Read-only secondary

" the documentation states

"In an auto-failover group

the secondary database can be used to offload read-only workloads." This directly supports using a replica for workloads like reporting to avoid impacting the primary.

2. Microsoft Learn

High availability for Azure SQL Database and SQL Managed Instance.

Reference: The "Locally redundant availability" section explains that high availability is a fundamental

built-in part of the service architecture

managed automatically

and is distinct from user-configured replicas for disaster recovery or read-scaling. This clarifies why option C is incorrect in the context of HA.

3. Microsoft Learn

Business continuity overview with Azure SQL Database.

Reference: This document defines Recovery Point Objective (RPO) and Recovery Time Objective (RTO). It clarifies that RPO is about data loss tolerance

which is a characteristic of the replication and backup technology

not the primary reason for implementing a replica. This shows why option D is an incorrect justification.

4. Microsoft Learn

Auditing for Azure SQL Managed Instance.

Reference: This page describes how to configure auditing directly on the server or database instance to write audit logs to storage accounts

Log Analytics

or Event Hubs. This process is independent of database replicas

invalidating option B.