Using .show queries to understand and optimize the KQL sent from PBI to Kusto
You use Kusto /ADX / KQL databases in Fabric with Power BI. You want to see the queries send to the backend from PBI and try to optimize the behavior. This article includes a query starting with the ".show queries" command that can help you identify the queries that need your attentionSearch text in a Kusto/ADX table using any number of terms from Power BI
In a table we have a free text column. Our goal is to return rows in which the text includes a list of terms which are pieces of text surrounded by special characters like comma, parentheses, space etc. We don't have such a filter in PBI so we'll use combination of a KQL operator - has_any and the mechanism of M dynamic parameters.Using Dual Mode dimension in Power BI with ADX data
Dual mode provides the best of using import mode and Direct Query mode for dimensions in PBI By using dual mode, you can reduce significantly the number of queries sent to ADX from PBI. The queries that involve the large tables using Direct Query will still be running entirely in ADX.Filtering and visualizing Kusto data in local time
Datetime values in Kusto (aka Azure Data Explorer/ADX/KQL database in Fabric) are assume to be in UTC time. In many cases users want to filter the data and see the results using their local time. This article compares different approaches to achieve this result and do it in an efficient way. The Kusto engine includes some new optimizations in this area but still caution is recommended. KQL examples that can be run on test data and are included also a paragraph and an example on how to implement in Power BICross highlight when using direct query in PBI
The default behavior of Power BI is that every visual can filter any other visual. This is fine when the model is using import mode and all queries are very fast and cheap. With Direct Query we need to be much more careful about the number of queries generated and send to the backend source
