3. 3 Azure Saturday 2018
About me
Developer and consultant for Microsoft BI at oh22data AG
Main Topics: SSAS, SSRS and MDS
>10 years of experience with database and BI development
Speaker at chapter meetings, national and international conferences, German PASS
local chapter lead, Data Platform MVP
@SQLMissSunshine
https://www.linkedin.com/in/gabimuenster/
https://www.xing.com/profile/Gabi_Muenster/cv
4. 4 Azure Saturday 2018
Agenda
What is Azure Analysis Services?
What is BI?
Which chances and risks do exist for BI in the cloud?
Which BI features are already available in the cloud?
How can Azure Analysis Services improve „BI in the cloud“?
5. 5 Azure Saturday 2018
What is Azure Analysis Services?
Enterprise grade OLAP engine and BI modeling platform
Fully managed platform-as-a-service (PaaS)
General availability since April 2017
Only Tabular Model mode so far
Multidimensional mode in consideration
Current Compatibility level: 1400 = SQL 2017
6. 6 Azure Saturday 2018
How to interact with Azure AS? (I)
You can use your standard on-prem toolset to interact with Azure AS
SSMS
SSDT
Team Foundation Services or Visual Studio Team Services
But you can also use new tools with new features:
Azure Analysis Web Designer
7. 7 Azure Saturday 2018
How to interact with Azure AS? (II)
Management:
Azure Portal
PowerShell
Azure functions
Rest API
Automization:
Azure Automation
SQL Agent
...
8. 8 Azure Saturday 2018
Why choose Azure AS?
Easy to get started
No waiting time for hardware or infrastructure
Azure AD makes start and collaboration easy
No additional tools or drivers need to be installed
Easy to move from current Tabular solutions on-prem into cloud
Deploy out of your existing environment
Backup/Restore
Easy to handle
Guided interaction with portal or „flexible“ interaction with PowerShell
Accessible from almost anywhere
10. 10 Azure Saturday 2018
What is BI?
“Business intelligence (BI) is an umbrella term that includes the applications,
infrastructure and tools, and best practices that enable access to and analysis of
information to improve and optimize decisions and performance.” (Gartner IT
Glossary)
“What is BI? There are two prevailing definitions out there – broad and narrow. The
broad definition … is that BI is a set of methodologies, processes, architectures, and
technologies that transform raw data into meaningful and useful information used to
enable more effective strategic, tactical, and operational insight and decision-
making. …the narrow definition is used when referring to just the top layers of the BI
architectural stack such as reporting, analytics and dashboards.” (Forrester Research)
11. 11 Azure Saturday 2018
What is BI?
Broad definition:
Load/ Enrich/Clean data => ETL
Transform data into readable model => ETL/DWH
Secure data => DWH/Mart
Analyze data => Mart/Datamining
Create szenarios, predictions,… => Datamining
Expose data => Presentation Layer
12. 12 Azure Saturday 2018
Corporate BI: Transformative data architecture
OLTP Apps Big Data & Analytics Comprehensive BI
Streaming
IoT
Social
Modern Data Warehouse
Big Data Storage
NoSQL
Data Lake
CRM
ERP
Business Analyst
BI Users
IT Pro
Data Mart or
Operational Data
Store
+ other data sources
LOB
Finance
Web
13. 13 Azure Saturday 2018
What is BI?
Special requirements
Combine data from different sources
Combine historical and current data
Bulk Load enabling
Read over write performance
Denormalization
Column Store Indexes
React to concurrency peaks
14. 14 Azure Saturday 2018
Which chances and risks do exist for BI in the cloud?
No „Microsoft only“ topic:
15. 15 Azure Saturday 2018
Which chances and risks do exist for BI in the cloud?
Chance Risk
Accessibility X
Scalability X
Improved cost management (PaaS, pay only when you need something) X (X)
Geo replication (support on roadmap) X
Security (X)
Online only (For BI a general risk, because on-prem BI very often
requires online access from customers) (X)
Self-service X (X)
16. 16 Azure Saturday 2018
Which BI features are already available in the cloud?
E(T)L
Azure Data Factory I + II (with SSIS)
Logic Apps
PowerBI Model/AAS Get Data Experience (using M)
Common Data Services/Common Data Services for Analytics (PowerBI)
...
DWH
Azure SQL Datawarehouse
Azure SQL Database
Azure Data Lake
...
17. 17 Azure Saturday 2018
Which BI features are already available in the cloud?
Mart
PowerBI Model (not accessible from other applications apart from Excel)
Common Data Services/Common Data Services for Analytics (PowerBI)
Datamining / Analysis
Azure Machine Learning
Cognitive Services
Azure Data Lake Analytics
…
Presentation Layer
PowerBI
18. 18 Azure Saturday 2018
Current Microsoft BI Landscape
Analysis
Server
Services
19. 19 Azure Saturday 2018
How can Azure AS improve „BI in the cloud“? (I)
Two questions:
1. What are the advantages compared to SSAS on-premise?
2. What are the advantages compared to already existing services for this feature
set?
20. 20 Azure Saturday 2018
Advantages compared to SSAS on-premise (I)
General cloud advantages
Easy to set up
Reduction of maintenance effort
Nearly no waiting times for supply of infratructure
Delete when no longer needed
Near high availability without additional cost
Earlier access to new features
21. 21 Azure Saturday 2018
Advantages compared to SSAS on-premise (II)
Scale up/down Scale out Pause/Resume
Usage Based Payment
22. 22 Azure Saturday 2018
Demo
Show Azure portal with Azure AS
General overview
Scale up/down
Scale out
23. 23 Azure Saturday 2018
How can Azure AS improve „BI in the cloud“? (II)
Azure Analysis Services offers the possibility to implement a semantic layer in the
cloud.
Even more, you have the chance to create a data virtualization layer!
From the feature perspective it covers the same elements that PowerBI does. So why
would you need Azure AS on top?
24. 24 Azure Saturday 2018
Advantages compared to PowerBI (I)
PowerBI Pro PowerBI Premium Azure Analysis Services
Model/Data Set Size
1 GB
not restricted
(initial upload restricted to 10 GB)
based on pricing tier (current max: 400
GB)
Refresh times per day
8 not restricted not restricted
Partitioning/Incremental refresh
No Public Preview Yes (not Basic)
Scalability
No Yes Yes
Cost management by pausing service
No (user licenses) No Yes
Usage of development toolset
PowerBI Desktop On the roadmap (June) Yes
Integrate into development strategies like
continuous integration
No On the roadmap (June) Yes
Extend modelling to functionality of
multidimensional
No No in evalution: vote here
25. 25 Azure Saturday 2018
Advantages compared to PowerBI (I)
Accessibility for different front-end tools
Usage based payment
Enterprise ready (only compared to Pro !)
Developer tool integration
Source code integration
Sizing and scalability (only compared to Pro !)
26. 26 Azure Saturday 2018
Examplatory cost comparison AAS vs PowerBI Premium
Business Scenario PowerBI Premium * AAS + PowerBI Pro
Dataset size 50GB
Business hours 24/7
100 Pro users, 400 consuming users
1 P1 Node x $ 4,995.00 ($ 4,995.00)
+ 100 x $ 9.99 ($ 999.00)
= $ 5,994.00
744 (31 days * 24 hours) x $ 4.06 ($ 3,020.64)
+ 500 x $ 9.99 ($ 4,995.00)
= $ 8,015.64
Dataset size 50GB
Business hours 10/5
100 Pro users, 400 consuming users
1 P1 Node x $ 4,995.00 ($ 4,995.00)
+ 100 x $ 9.99 ($ 999.00)
= $ 5,994.00
220 (22 days * 10 hours) x $ 4.06 ($ 893,20)
+ 500 x $ 9.99 ($ 4,995.00)
= $ 5,888.20
Dataset size 50GB
Business hours 24/7
500 Pro users
1 P1 Node x $ 4,995.00 ($ 4,995.00)
+ 500 x $ 9.99 ($ 4,995.00)
= $ 9,990.00
744 (31 days * 24 hours) x $ 4.06 ($ 3,020.64)
+ 500 x $ 9.99 ($ 4,995.00)
= $ 8,015.64
* Source: https://powerbi.microsoft.com/en-us/calculator/
27. 27 Azure Saturday 2018
Close the gap between user requests and Enterprise BI
Key User Self Service
BI
AAS Web Designer
+
Enterprise
Migration
Enterprise BI Frontends Standard
Users
28. 28 Azure Saturday 2018
Demo
• Show Analysis Services Web Designer
• Import PowerBI Model
• Show BISM Normalizer
• Compare new AAS model to „enterprise model“
29. 29 Azure Saturday 2018
Conclusion
• Does AAS add considerable value to „BI in the cloud“?
Definitely yes!
• Is AAS a must-have service in all scenarios?
Not necessarily! PowerBI Premium grows to be a suitable alternative in
some cases a better price, but you won‘t be able to use other
visualization tools on it.
• In which scenarios does AAS make sense?
Create a semantic layer/data virtualization layer as single point of
truth for your reporting and analytics
Migration existing semantic layer (SSAS Tabular) into the cloud
Integrate PowerBI Pro developments into existing semantic layer
30. 30 Azure Saturday 2018
References
Azure Analysis Services
https://docs.microsoft.com/en-us/azure/analysis-services/analysis-services-overview#
https://www.kasperonbi.com/analysis-services-in-azure-when-and-why/
https://www.kasperonbi.com/getting-your-on-premise-ssas-tabular-model-to-azure/
https://azure.microsoft.com/de-de/blog/1400-models-in-azure-as/
What is BI?
http://www.gartner.com/it-glossary/business-intelligence-bi/
http://blogs.forrester.com/boris_evelson/10-04-29-
want_know_what_forresters_lead_data_analysts_are_thinking_about_bi_and_data_domain
BI in the cloud
http://biinthecloud.com/
https://cloud.oracle.com/business_intelligence
https://www.microsoft.com/en-us/sql-server/business-intelligence