ARM templates are a way of defining your infrastructure using declaritive statements, as opposed to scripts. You define what you want your resources to look like, including the properties and values, and Azure will ensure they look that way. Scripts suffer from the problem of not being able to run the same script twice, whereas ARM templates can be deployed again and again without issue. This is sometimes called Desired State Configuration (or DSC). Refer to Microsoft Doc:

https://docs.microsoft.com/en-us/learn/modules/explore-provision-deploy-non-relational-data-services-azure/2-describe-provision-non-relational-data-services

Graph databases allow you to easily navigate the relationships between the data (edges) and not just the data itself. It was specially designed for things like social networks. Refer to Microsoft Doc:

https://docs.microsoft.com/en-gb/azure/architecture/data-guide/big-data/non-relational-data

ETL means that you transform the data before it gets loaded into the data store, which can allow you to filter out sensitive fiels and personally-identifiable information.

You can select from any of the still-supported versions of SQL Sever on a VM.

Azure Data Lake is the correct answer. Blob Storage is also a place to store data before processing, but if you can't estimate how much data, there are limits on the size per account. Cosmos DB is not great for large amounts of unprocessed data. A Data Lake is a much better answer for ingesting data before processing. Synapse is a good data warehouse solution, but typically is used after you have ingested and processed the data. Refer to Microsoft Doc:

https://docs.microsoft.com/en-ca/azure/storage/blobs/data-lake-storage-introduction

Analytics workloads focus on running reports, with complex and long-running queries. They are best unnormalized or at least partially-unnormalized. Data that can change at any moment would make such a DB difficult as reports would have different results between runs. Refer to Microsoft Doc:

https://docs.microsoft.com/en-us/azure/architecture/data-guide/relational-data/online-transaction-processing

Treemaps are charts of colored rectangles, with size representing the relative value of each item. They can be hierarchical, with rectangles nested within the main rectangles.

The SQL syntax requires the verb SELECT, followed by a list of column names or *, followed by FROM, followed by the table name, and ending in a semi-colon.

Indexes allow a database query engine to find data much faster than searching the entire table for the data.

Batch processing is better designed to handle large quantities of data efficiently. Stream processing is intended for small batches of data or individual items.