Business Intelligence Solution Components Overview

Siddharth Mehta
A free video tutorial from Siddharth Mehta
Enterprise Cloud Architect, Published Author, Cloud Geek
4.2 instructor rating • 11 courses • 47,361 students

Lecture description

This lecture provides an overview of BI architecture components like Source Systems, Staging Zone, Landing Zone, Data Warehouse, Data Mart, Reports and Dashboards

Learn more from the full course

SQL Server Analysis Services - SSAS, Data Mining & Analytics

Develop SSAS cubes from data warehouse on Multidimensional & Tabular modes with Dimensional & Data Mining Models

07:33:25 of on-demand video • Updated March 2019

  • Downloadable Course Content - FREE SQL Server 2016 EBook, 150+ Interview Questions, 100+ curated reference material links, 70+ MDX / DAX Queries, Links to 1 Sample OLTP Database and 1 Sample Data Warehouse with sample data - schema diagrams - documentation - dimensional modeling workbook, 1 Sample Multidimensional OLAP Database + 1 Sample Tabular OLAP Database containing fully functional cube with sample data, and Full project developed in the course.
  • By end of this course, you should be able to work with actual clients, technology architects, or team leads on a large scale SQL Server Analysis Services / MDX / Data Mining Project
  • Learn SQL Server Analysis Services ( SSAS ) Multi-dimensional as well as Tabular, Multidimensional Data Expressions ( MDX ) query language, Data Analysis Expressions (DAX) expression language and Data Mining on latest version of SQL Server - 2016 on a fast track with more practicals and limited theory necessary for fundamentals.
  • Develop Data Mining Models on algorithms like Decision Trees, Naive Bayes, Clustering, Neural Network, Time Series, Linear Regression and Sequence Clustering
  • Test your knowledge with more than 150 SSAS, MDX, DAX and Data Mining interview or Microsoft BI certification practice test questions. Every few weeks, new interview questions are added to the course.
  • Design, Model and Build SSAS database, data warehouse or data mart, and database objects like facts, dimensions, cubes, measures, perspectives, etc
  • Deploy SSAS solutions and learn SSAS administration tasks like backup and restore, processing dimensions and measures, building aggregations, etc
  • Develop MDX and DAX code with confidence
  • Develop SSRS Reports using MDX as query language and SSAS as data source
  • Debug issues and performance tune SSAS solutions for optimal performance
  • Learn about Microsoft Business Intelligence Architecture, and understand how SSAS fits in this architecture along with other tools like SSIS, SSRS, Master Data Services ( MDS ), Data Quality Services ( DQS ), Powerpivot, PowerView, SQL Server Database Engine, SSMS, SSDT and other tools.
English [Auto] Hello and welcome back in this chapter will be looking at a solution architecture or you. So let's start with some background. If you look at this diagram those who are seen architectural diagrams earlier will be able to easily say that this does not look like an architectural diagram. Well it's a simplified version of architecture diagram to easily understand what an architect looks like. So before we start discussing this diagram let me say just one thing. If you are asked to draw an architecture diagram Please do not draw these kind of background. This diagram is simplified so that beginners can easily grasp photo the different components that goes into a solution architecture. Typically architecture like it has all those porticoes and horizontals like logging security than their dad makes and models that are offered and on the front of a deal how all the reporting doors and the cons you must layaways this diagram is just a simplified version of it. So now let's take a minute to understand what goes into a typical Endore and be a solution architecture. Is this better off whether it's Microsoft or article or any MBA architecture. This is more or less a broader view of the business into a just solution. So now let's look at the what are the typical needed high level components. Take a look at today at Give a thought to what is the first thing that you see on the left and you'll see a different kind of source systems. Here we have to present it mostly all deep in source systems. Now there can be different kind of other sources TMS also like Excel flat file or. Other days we have non-date big data and a lot of different systems. The only way to solve the problem of big data solution that easily into an election system but still for the purpose of considering datasource is if you see it starts with this source systems so that this other first component. Now for the timing Barkworth the CDR. Let's take the second part of it. I know a lot of source systems that would help would be sourced from the systems collected from that because there's that layer of transaction systems. These are not meant for me to know that in all the systems because they are done with performance issues there can be issues on authority that can be issues on scalability. So generally whatever data that is to be analyzed that is to be reported. That is taken out into a separate system. So what we do when we expect data and stage Axum that means we parted Sumba. So that comes to staging area. Now we also call it landing area or the difference between staging and landing. The difference is that once the date they staged it might be just an exact replica are just a snapshot or a copy of the data that we are expecting. But we may want to Glenn say if we want to play some business logic we won't grow some transformation. We just want to get it into the shape that is suitable for the business. That is generally what is done between staging and landing. So from a higher level both falls into the same zone. It can be implemented in the phone of course a database is as simple a stack but that is where the staging and lending area comes. All of us were how hard the beat the White House at some point in time somewhere. Now there can be many many many different source systems into an organisation and that can with different indignations system data flows from one end to another. But we need a single version of that too. The nation needs a single version of the truth that anyone needs any kind of beat. They can just go to a warehouse get their snapshot copy a slice of it and then use it in their own small scale ordinance in a specialized some systems are suited for the business. This is why the data warehouse at all. Now it seems that we are collecting a lot of data now in order for us to make this data transfer. What we need is an ETF. What is median median is extract transform loss. So we need a driving mechanism that can drive data from let's say from the sources to the landing or staging area from the landing or ceiling. We may again in the scheme of the new Dabb because we need a separate kind of schema for it at ALS. So this is where it all comes in the picture generally in Michael's of B.A technology stack. This e-tail will be seekers will integration services. This technology will build back. I just read that it'll contain the logic to the data from source systems transform it as required and load it into the destination systems. It may be a landing area or maybe a bad house but it may well be tomorrow. Now here comes the data on Mark. Mark. As we said earlier data from data warehouses whatever systems that may need data they will create a small sub copy of their data suited for the requirements of their own applications. For example there is an application for a job that is an application for a finance that counts are different the kind of applications enterprise may require so that they will create market specialized for those particular areas. Take any one model let's say just for finance. This all the finance and operations and those kind of data into that. Generally in terms of Microsoft we say architecture that it would help you will come to work skilled in the lead of chapters. But for now it is to do understand it in a simplified manner. We can say that a cube is multi dimensional database laking seekers at all we have a database of course data. This is an analytical database and all database online and processing Vitale's. So this is where the data lies. This can be incept process segregate and transform the copy of the data that mostly will come from Andy though it helps. And then finally once for you how all this data you will see this Dagget you can easily make code that is reporting clear reporting it can be from any number of moderates or any number of sources. It need not just be a tomorrow. It can be from oil DP sources it can lead from warehouses or indeed from Mars. But in our solution because we want takes and not a detailed 86 here in the cube. So reporting all three of marks. And this is very important dashboards them and dashboards contain different kind of reboarding elements like KBI said at a school like ours. There are geographical reports containing maps to all those different kinds of reports out there. This and over and on a solution that components. I will see you in the next step of.