Ssis sql ssrs_ssas_sp_mdx_hb_li

Hong-Bing Li

[email protected]

Slide Contents

3 SQL Server Integration Services (SSIS)

8 SQL Programming

15 SQL Server Reporting Services (SSRS)

20 SQL Server Analysis Services (SSAS)

27 SharePoint Server 30 MDX Programming

This portfolio contains examples of my business intelligence

projects using Microsoft BI Product Stack.

SQL Server Integration Services (SSIS)

3

A new database “All Works” is setup as the staging area for the ETL process. A thorough understanding of the relationships between the tables in the following data diagram is important in determining the sequence of tables to be loaded and in enforcing referential integrity.

4

A Script Task is utilized to maintain multiple sets of variables with scripts in C#, for instance, one for keeping track of row counts of data processed dynamically at the

folder level, one for row counts at the file level.

One SSIS package is created to do ETL for one target table. The following illustrates the data processing within the Job Timesheets package: the data process pipeline starts by extracting data from a CSV file. The data is then conversed, processed and transformed (filter, remove duplicates, lookups, validate) as it passes through the pipeline, and is finally loaded into the target job timesheets table either as inserts or updates. It logs any rows that error out for review and correction. Similarly, seven more packages are generated for seven target tables.

6

7

A Sequence Container is deployed to run the eight ETL packages in sequence based on the relationships between the tables in the “All Works” Database to ensure referential integrity. If the eight packages are processed successfully, data maintenance tasks are performed. A success or failure notice email will be sent out depending on whether the data maintenance tasks are all successfully completed or not.A Master Package is created to contain the Sequence Container, the maintenance tasks and the email notices; then a SQL Server Agent Job is setup to run the Master Package on a predefined schedule to automate the entire data processing procedure.

SQL Programming

SQL Programming covers a broad range of content: Data Types and

Variables, Specialized Functions, Stored Procedures, Control-Of-Flow Operations,

Error Handling, Etc. Due to time/space constraints, I will illustrate only some of

the relatively new SQL functions/procedures/ programs that I have used in my

projects in the following:

1. Data Modification: OUTPUT clause, MERGE, select TOP(with ties) option

2. Aggregation: RANK() OVER (Partition by), GROUPING SETS

3. SQL Functions: PIVOT, CROSS APPLY

4. Common Table Expression (CTE), Correlated and Recursive Queries

5. User Defined Functions (UDF) that returns TABLE type

6. Error Handling Program:

Begin Try...End Try

Begin Catch...End Catch

SQL Server Reporting Services (SSRS)

15

16

17

Continued: Primary Dashboard

18

19

Whenever users make a selection on the "City" parameter, the cascading parameter "Product SKU" is processed immediately. Its values are filtered dynamically based on two factors:

A. Selected cities B. Product SKU with dollar Sales greater than “0 “

The technique to implement cascading parameters in SSRS using MDX, which is based on OLAP, is somewhat more complex than that using SQL, which is based on regular OLTP RDBMS.

SQL Server Analysis Services (SSAS)

20

21

The Development and Deployment of the All Works SSAS Cube Data Source View of the Star Schema

22

Browsing the All Works Cube Data

23

Definition of Calculated Members

24

Definition of Key Performance

Indicators (KPIs)

25

This is an example of a KPI developed from the AllWorks.cube, which is deployed to the Excel Spreadsheet for the end users. All the data in the cube, including all KPIs,

can be explored through the Pivot Table Field List.

Partitions Performed for the “All Works” Cube

SharePoint Server

27

Employee Labor Report Deployed to a SharePoint Server

configure security settings: give users/

groups appropriate permissions.

MDX Programming

30

MDX query for the primary dashboard (Slide 15 and 16 in SSRS Section)

This MDX Query shows two measures, Internet Sales Amount and Internet Freight Cost, are sliced against three dimensions:

Date (FY Year), Product and Customer (Country)

---The End---

Hong-Bing Li, February, 2015

Thank you for viewing this presentation!