Examkin

Question

Difficulty: HardData Integration and Analytical Storage

A financial services company is designing a new analytical data platform on Azure to query historical audit logs. The audit logs are stored in Parquet format in an Azure Data Lake Storage Gen2 account.

The platform must meet the following requirements:
- Volume: 80 TB80\text{ TB} of data is currently stored, growing by 2 TB2\text{ TB} per month.
- Velocity: Batch uploads occur once every week.
- Query Latency and Pattern: Data analysts run highly complex, ad-hoc exploratory queries to investigate anomalies. These queries occur irregularly, and the system may remain idle for several days.
- Resilience: The data must remain available even during a complete regional datacenter outage.
- Security: Access to the data lake must be delegated to external audit teams via Shared Access Signatures (SAS) that can be immediately revoked if security policies are violated.

You need to recommend the storage configuration and the query engine for the platform.

Which two components should you include in the recommendation? (Select TWO.)

  1. An Azure Synapse Analytics serverless SQL pool to query the Parquet files directly in the data lake.Answer
  2. A storage account configured for Geo-zone-redundant storage (GZRS) with access granted via Shared Access Signatures (SAS) that reference a stored access policy.Answer
  3. C
    An Azure Synapse Analytics dedicated SQL pool to ingest the Parquet files into relational tables before running queries.
  4. D
    A storage account configured for Locally redundant storage (LRS) with access granted via Shared Access Signatures (SAS) that reference a stored access policy.
  5. E
    A storage account configured for Geo-zone-redundant storage (GZRS) with access granted via ad-hoc Shared Access Signatures (SAS) configured with a 3-year3\text{-year} expiration date.

Answer

The platform should include an Azure Synapse Analytics serverless SQL pool to execute ad-hoc queries, and a storage account configured for Geo-zone-redundant storage (GZRS) with access granted via Shared Access Signatures (SAS) referencing a stored access policy.
The correct recommendation consists of using an Azure Synapse Analytics serverless SQL pool and configuring the storage account for Geo-zone-redundant storage (GZRS) using Shared Access Signatures (SAS) associated with a stored access policy. A serverless SQL pool is optimal for ad-hoc, infrequent queries on data lake files because you only pay per query, avoiding idle costs. GZRS provides resilience against both zone and regional failures by replicating data across zones and to a secondary region. Associating the SAS with a stored access policy allows the administrator to revoke access immediately by modifying or deleting the policy.

Step-by-Step Solution

1
Analyze the query pattern to choose the optimal query engine.
Serverless SQL pools are selected because the queries are ad-hoc, irregular, and the platform remains idle for days. This eliminates the persistent costs associated with running a dedicated SQL pool cluster when idle.
Dedicated SQL pools run continuously and incur significant costs even when idle, making them unsuitable for sporadic exploratory query workloads.
2
Evaluate the storage redundancy requirement to ensure availability during regional outages.
Geo-zone-redundant storage (GZRS) is selected. GZRS replicates data across three availability zones in the primary region and to a secondary region, providing protection against a regional disaster.
Locally redundant storage only replicates data within a single datacenter and cannot survive a regional outage.
3
Determine the secure access method that allows immediate revocation.
Shared Access Signatures (SAS) referencing a stored access policy are chosen over ad-hoc SAS tokens with long lifetimes.
Ad-hoc SAS tokens cannot be revoked individually without rotating the storage account keys. Referencing a stored access policy allows immediate revocation by deleting or modifying the policy.

Key Concept

Selecting the optimal query compute engine, storage redundancy, and access delegation mechanisms in Azure to balance cost, resilience, and security.
Rate this question

Topics

Data Integration and Analytical Storage

More questions from this topic