DP-900 Practice Question: Identify considerations for relational data on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Provisioned DTU

  Why it's wrong here

  Provisioned DTU is wrong because the DTU-based purchasing model bundles compute, storage, and I/O into fixed service tiers (Basic, Standard, Premium) that are billed continuously per hour, regardless of actual database usage. A DTU represents a blended measure of CPU, memory, and I/O, but the key flaw here is that it cannot auto-pause or scale down to zero during idle times. An intermittent reporting workload would therefore incur full charges even when no reports are being run, making it economically inefficient compared to serverless.

  When this WOULD be correct

  For a predictable workload with consistent usage patterns where you need guaranteed performance and can commit to a fixed resource level, such as a production OLTP system with steady traffic.

• ✗

  Provisioned vCore

  Why it's wrong here

  Provisioned vCore is wrong for this scenario because it charges a flat hourly rate for a fixed number of virtual cores, memory, and storage, and it lacks any built-in auto-pause mechanism. You can manually pause the database to avoid charges, but that requires operational intervention and is not automated, unlike the serverless tier's automatic behavior. An intermittent reporting application would incur continuous billing unless someone manually manages the pause/resume lifecycle, making it less cost-effective than serverless.

  When this WOULD be correct

  A question where the database requires predictable performance with reserved capacity, such as a production OLTP workload with consistent traffic, and the organization needs to isolate compute from storage for scaling or licensing flexibility.

• ✓

  Serverless

  Why this is correct

  Serverless is the correct choice because Azure SQL Database's serverless compute tier is explicitly designed for intermittent workloads like this reporting application. It automatically pauses the database after a configurable period of inactivity (default 60 minutes) and resumes instantly on the next connection, so compute is billed only on actual usage per second during active periods. Storage remains billed separately at all times, but eliminating idle compute charges makes it far cheaper than provisioned models for a reporting app that is used sporadically.

• ✗

  Hyperscale

  Why it's wrong here

  Hyperscale is incorrect because it is a service tier optimized for very large databases that need rapid storage scaling and high-throughput continuous workloads, not for infrequent, short-lived reporting queries. Although Hyperscale separates compute and storage, its compute nodes must remain active to serve queries across many replicas and do not support an auto-pause feature. Running a small, intermittent reporting app on Hyperscale would waste resources and money, as you'd pay for always-on compute capacity that remains idle most of the time.

  When this WOULD be correct

  A question where the database requires massive storage (e.g., over 4 TB), needs rapid scaling for unpredictable high loads, or demands fast recovery from failures. For example: 'A company has a 10 TB database with sudden spikes in traffic and needs near-instant recovery. Which service tier should they choose?'

Option-by-option analysis

Why each answer is right or wrong

Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The DP-900 exam frequently reuses these exact scenarios with slightly different constraints.

Analysis generated from the official DP-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”