Recommended Data Store: Azure Blob Storage

C

Object storage is optimized for storing and retrieving large binary objects (images, files, video and audio streams, large application data objects and documents, virtual machine disk images). Large data files are also popularly used in this model, for example, delimiter file (CSV), parquet, and ORC. Object stores can manage extremely large amounts of unstructured data.

Based on the given requirements, the most appropriate non-relational data store to recommend is an object store.

An object store is a type of non-relational data store that is optimized for storing and retrieving large unstructured data sets, such as text files, videos, audio streams, and virtual disk images. It is also designed to handle data of varying sizes, making it ideal for storing large files.

Object stores use a flat namespace, meaning that each object is identified by a unique key or identifier. This satisfies the requirement of storing a unique ID for each file. Additionally, object stores can store metadata associated with each object, such as creation date, file size, and file format, which satisfies the requirement to store some metadata.

The other options provided in the question are not as well suited for the given requirements:

• Key/value stores are optimized for storing and retrieving simple key/value pairs and may not be well suited for storing large unstructured data sets.
• Columnar stores are optimized for storing and retrieving large sets of structured data, such as tables with many columns, and are not ideal for unstructured data.
• Document stores are optimized for storing and retrieving semi-structured data in the form of JSON, XML, or BSON documents and may not be well suited for storing large unstructured data sets.

Therefore, the recommended non-relational data store for this scenario is an object store.