Date & Time granularity

Dremio gives you control of the granularity of Date & Time values as dimensions in aggregate reflections.  This control influences both the behavior and size of the resulting reflection.

By default, Dremio will store the Date & Time value “rolled up” to (D) day granularity
As an option, you can store Date & Time values ( Read more... )

Dimensions and measures
Dimensions are used when grouping records, such as in GROUP BY queries. Dimensions are things like:

• Location
• Department
• size

Measures are numeric values computed by queries, such as

• cost
• distance

Aggregate reflections
Aggregate reflections greatly improve the performance of aggregating queries such as GROUP BY, SUM, AVG, etc. They optimize distinct value expressions and optionally store an approximate count distinct on a column.

When you create an aggregate reflection, Dremio makes recommendations for the dimension and measures in the ( Read more... )

Raw reflections
Raw reflections preserve the row-level detail of the dataset and therefore support needle-in-a-haystack queries that require this level of detail.
They provide several benefits, including:
- Your choice of columns in each reflection to target each reflection.  Removing unused columns reduces input scanning and resulting query time
- ( Read more... )

https://university.dremio.com/courses/course-v1:dremio+D102+2019/about

Reflections are optimized data structures that can dramatically accelerate query execution.

Reflections are physically optimized representations of source data. When they ( Read more... )

Type conversions
Dremio performs type conversions quickly without requiring knowledge of SQL
Dremio provides easy and powerful type conversion in a few mouse clicks.  This capability enables data analysts and architects to successfully prepare complicated datasets for advanced analysis without engaging complicated ETL scripts, conversion tools, or ( Read more... )

( Read more... )

The key words are:
- Sources (https://university.dremio.com/courses/course-v1:dremio+D101+2018D101/courseware/f038f3b2912d4dc6ab6e154cc824780a/ ( Read more... )

https://finacademy.net/materials/article/power-bi

For over a decade, SQL Server programmers had to remember: avoid using table variables with lathe count of rows in them. The reason was their low performance in that case:
making an execution plan, SQL Server estimated a table variable as always having 1 row (even in fact the the table variable had millions of rows).
In SQL Server 2019 an ( Read more... )