Netherlands: Software

Introductie van Micorosoft SQL Server 2016

Issue link: http://hub-nl.insight.com/i/692679

Contents of this Issue

Navigation

Page 205 of 212

193 C H A P T E R 9 | Introducing Azure SQL Data Warehouse statistics depends on the characteristics of your data. As you develop your plan, consider the following guidelines: Update at least one statistics object in each table containing data to update the row and page count information automatically as part of the same operation. Regularly review columns that appear often in JOIN, GROUP BY, ORDER BY, and DISTINCT clauses. An ascending key column for which new values are added frequently, such as a transaction date column, likely requires more frequent updates in order to include these new values in the statistics histogram. If a column has a set of values that remain static, reduce the frequency of updates or eliminate updates altogether. Because each statistics object is updated in a series, avoid using UPDATE STATISTICS for wide tables. Instead, consider using UPDATE STATISTICS (STATISTIC_OBJECT) to update specific objects as needed. Integration options After you load data into SQL Data Warehouse, your users can use a variety of tools to gain deeper insights into that data. SQL Data Warehouse integrates easily with other components of the Microsoft analytics and business-intelligence stack. In this section, we take a closer look at integrating with the following cloud-based tools: Azure Machine Learning Azure Stream Analytics Power BI Note Considering the volume of data stored in a typical data warehouse and the higher volume of data recommended for SQL Data Warehouse, the use of cloud-based analysis tools helps you avoid both data egress charges and latency associated with bringing data back to on-premises installations. Predictive modeling in Azure Machine Learning Azure Machine Learning is a managed analytics service that you can use to build complex predictive models. You can use SQL Data Warehouse as both a source and a destination for your machine- learning models. To retrieve data from SQL Data Warehouse for your experiment, expand the Data Input and Output module category and then drag the Reader module to the experiment canvas. In the Properties window, shown in Figure 9-8, select Azure SQL Database in the Data Source drop-down list and then provide the following connection details in the respective text boxes: Database Server Name, Database Name, Server User Account Name, Server User Account Password, and Database Query. You have the option to select the Accent Any Server Certificate (Insecure) check box if you want to omit the step to review the site certificate before accessing your data, although this action is not recommended. The Properties window also includes the Use Cached Results check box that you can select if you want to read the data once and then use the cached results for each subsequent run of your experiment.

Articles in this issue

Archives of this issue

view archives of Netherlands: Software - Introductie van Micorosoft SQL Server 2016