Migrating Enterprise BI to Azure
Best Practices
Wlodek Bielski
SQLSat Kyiv Team
Eugene Polonichko
Oksana Tkach
Oksana Borysenko
Denis Reznik
Mykola Pobyivovk
Yevhen Nedashkivskyi
Sponsors
Session will begin very soon :)
▪ Please complete the evaluation form from
your pocket after the session. Your feedback
will help us to improve future conferences
and speakers will appreciate your feedback!
▪ Enjoy the conference!
▪ MCSD: Azure Solutions Architect
▪ MCSE: Cloud Platform and Infrastructure
▪ MCSE: Data Management and Analytics
▪ AWS Certified Solution Architect
▪ Google Professional Data Architect
▪ Google Professional Data Engineer
About me
Agenda
▪ Setting up Azure environment
▪ Networking, VPN and on-prem connectivity
▪ Migrating Analysis Services
▪ Migrating Staging / DWH
▪ Migrating ETL/ELT
▪ Enabling Big Data scenarios
SETTING UP AZURE
ENVIRONMENT
Azure purchasing options
▪ Free Trial
▪ Pay-as-you-Go
▪ Open Licensing
▪ Enterprise Agreement
▪ Cloud Solution Provider (CSP)
Dev/Test offering
Setting up Azure environment
▪ Dev/Test + Prod subscriptions
▪ Design multiple Resource Groups
▪ bi-prod
▪ bi-test
▪ bi-dev
▪ bi-common
▪ Networking (VNet, VPN)
▪ Might be costly to change later
NETWORKING
AND CONNECTIVITY
Networking part
▪ Point-to-Site VPN – quick start
▪ Site-to-Site VPN – will take time
▪ ADF v2 – Self-Hosted IR
▪ On-premises Gateway
(Analysis Services and Power BI)
VPN: Hub and Spoke topology
MIGRATION JOURNEY
Migration Roadmap
Starting
point
SQL 2008
R2
Integration
Services
Analysis
Services
Reporting
Services
Phase
1
SQL
Server
IaaS
Integration
Services
IaaS
Analysis
Services
IaaS
Reporting
Services
IaaS
Phase
2
SQL
Server
IaaS
Azure Data
Factory v2
with SSIS
Azure
Analysis
Services
Power BI
Phase
3
Azure SQL
DB / DWH
Azure Data
Factory v2
Azure
Analysis
Services
Power BI
Database ETL/ELT OLAP Reporting
Phase 1: Lift & Shift
Starting
point
SQL 2008
R2
Integration
Services
Analysis
Services
Reporting
Services
Phase
1
SQL
Server
IaaS
Integration
Services
IaaS
Analysis
Services
IaaS
Reporting
Services
IaaS
Phase
2
SQL
Server
IaaS
Azure Data
Factory v2
with SSIS
Azure
Analysis
Services
Power BI
Phase
3
Azure SQL
DB / DWH
Azure Data
Factory v2
Azure
Analysis
Services
Power BI
Database ETL/ELT OLAP Reporting
Main objectives
▪ Migrate solution to VM(s)
▪ Take advantage of Azure IaaS services
▪ Setup secure and manageable environment
▪ Gain confidence
Lift & Shift options
▪ Backup-restore
▪ Data Migration Assistant
▪ IaaS – Windows template
▪ IaaS – SQL Server templates
▪ SQL Database Managed Instances
▪ License considerations
Data Migration Tools
▪ Data Migration Assistant
▪ Supports both scenarios
▪ On-premises -> IaaS -> PaaS
▪ Azure Database Migration Service
▪ Managed service for migrations at scale
PaaS services coverage
Covered
▪ Database Engine
(SMP / MPP)
▪ Analysis Services
▪ Integration Services
(via ADF)
Not covered
▪ SQL Server Agent
▪ Reporting Services
▪ Master Data Services
▪ Data Quality Services
…Managed Instance?
SQL Database Managed Instance
Infrastructure recommendations
▪ Plan for PaaS from the beginning
▪ Decouple services to separate VMs
▪ Use Premium managed disks
▪ Less burden with storage accounts (leftovers)
▪ Consider 2/4 TB disks (P40/P50)
▪ Azure Backup supports up to 16 data disks
▪ Azure Backup + Recovery Vault
▪ Use both (VM + disks/data)
Licensing considerations
▪ Pay-as-you-Go vs BYOL
▪ BYOL: 2008 R2+ are supported
▪ Consider SQL Server Developer Edition
▪ Consider memory-optimized VMs (G- / M-)
for Analysis Services Tabular
▪ Dev/Test subscriptions
Cost control recommendations
▪ Start small ☺
▪ Cloudyn – Azure Cost Management
▪ Consider Auto-Shutdown
▪ Reserved instance? Commitment vs flexibility
▪ DEMO – Azure Portal
Phase 2: Transition
Starting
point
SQL 2008
R2
Integration
Services
Analysis
Services
Reporting
Services
Phase
1
SQL
Server
IaaS
Integration
Services
IaaS
Analysis
Services
IaaS
Reporting
Services
IaaS
Phase
2
SQL
Server
IaaS
Azure Data
Factory v2
with SSIS
Azure
Analysis
Services
Power BI
Phase
3
Azure SQL
DB / DWH
Azure Data
Factory v2
Azure
Analysis
Services
Power BI
Database ETL/ELT OLAP Reporting
Main objectives
▪ Start using PaaS services
▪ Optimize IaaS infrastructure
▪ Understand and control costs
PaaS recommendations
▪ Start with Analysis Services – quick win
▪ Consider Azure SQL DB for DWHs < 1 TB
▪ Redesign ETL/ELT in phases
▪ Start with SSIS runtime
▪ Rewrite to pure ELT afterwards
▪ Orchestration instead of linear pipelines
Analysis Services deployment options
Multidimensional
▪ On-premises
▪ IaaS (BYOL)
▪ IaaS (SQL Server template)
Tabular
▪ On-premises
▪ IaaS (BYOL)
▪ IaaS (SQL Server template)
▪ Azure Analysis Services
▪ Power BI Premium
(de facto)
Azure Analysis Services
▪ Scaling up / out – finally!
▪ Pricing tiers
▪ Replicas
▪ Managed backups
▪ Expensive service – use with care
▪ REST API
▪ Can be automated via Logic Apps / Functions
SSAS: On-premises Data Gateway
Azure Data Factory v2
▪ Great orchestration tool with simple GUI
▪ Integrated Git support – easy versioning
▪ Runs SSIS (lots of other engines as well!)
▪ Supports ARM templates
▪ Not an ETL tool actually!
Power BI vs SSRS
▪ Power BI has limited support for pinning
SSRS reports
▪ Paginated reports likely to be supported
in future Power BI releases
▪ Plan for leveraging Power BI
from moment of IaaS migration
▪ Keep SSRS on small separate VM if needed
Phase 3: Cloud Native
Starting
point
SQL 2008
R2
Integration
Services
Analysis
Services
Reporting
Services
Phase
1
SQL
Server
IaaS
Integration
Services
IaaS
Analysis
Services
IaaS
Reporting
Services
IaaS
Phase
2
SQL
Server
IaaS
Azure Data
Factory v2
with SSIS
Azure
Analysis
Services
Power BI
Phase
3
Azure SQL
DB / DWH
Azure Data
Factory v2
Azure
Analysis
Services
Power BI
Database ETL/ELT OLAP Reporting
Main objectives
▪ Operate pure PaaS solution
▪ Get ready for Azure DWH with ELT
▪ Enable Big Data / real-time scenarios
Main activities
▪ Migrating to Azure SQL Database
▪ Getting rid of SSIS / ETL
▪ Rewriting new ELT orchestrations
▪ Minimizing VM workloads
Beyond classical DWH/BI
▪ HDInsight
▪ Data Lake Storage / Analytics
▪ Databricks
▪ .. Data Factory to orchestrate them all!
Wrap-up
▪ Phased approach
▪ Organization needs to learn new ways
▪ Not too fast..
▪ ..but it’s just the right time to start!
Thank you!
Sponsors

Migrating Enterprise BI to Azure

  • 1.
    Migrating Enterprise BIto Azure Best Practices Wlodek Bielski
  • 2.
    SQLSat Kyiv Team EugenePolonichko Oksana Tkach Oksana Borysenko Denis Reznik Mykola Pobyivovk Yevhen Nedashkivskyi
  • 3.
  • 4.
    Session will beginvery soon :) ▪ Please complete the evaluation form from your pocket after the session. Your feedback will help us to improve future conferences and speakers will appreciate your feedback! ▪ Enjoy the conference!
  • 5.
    ▪ MCSD: AzureSolutions Architect ▪ MCSE: Cloud Platform and Infrastructure ▪ MCSE: Data Management and Analytics ▪ AWS Certified Solution Architect ▪ Google Professional Data Architect ▪ Google Professional Data Engineer About me
  • 6.
    Agenda ▪ Setting upAzure environment ▪ Networking, VPN and on-prem connectivity ▪ Migrating Analysis Services ▪ Migrating Staging / DWH ▪ Migrating ETL/ELT ▪ Enabling Big Data scenarios
  • 7.
  • 8.
    Azure purchasing options ▪Free Trial ▪ Pay-as-you-Go ▪ Open Licensing ▪ Enterprise Agreement ▪ Cloud Solution Provider (CSP)
  • 9.
  • 10.
    Setting up Azureenvironment ▪ Dev/Test + Prod subscriptions ▪ Design multiple Resource Groups ▪ bi-prod ▪ bi-test ▪ bi-dev ▪ bi-common ▪ Networking (VNet, VPN) ▪ Might be costly to change later
  • 11.
  • 12.
    Networking part ▪ Point-to-SiteVPN – quick start ▪ Site-to-Site VPN – will take time ▪ ADF v2 – Self-Hosted IR ▪ On-premises Gateway (Analysis Services and Power BI)
  • 13.
    VPN: Hub andSpoke topology
  • 14.
  • 15.
    Migration Roadmap Starting point SQL 2008 R2 Integration Services Analysis Services Reporting Services Phase 1 SQL Server IaaS Integration Services IaaS Analysis Services IaaS Reporting Services IaaS Phase 2 SQL Server IaaS AzureData Factory v2 with SSIS Azure Analysis Services Power BI Phase 3 Azure SQL DB / DWH Azure Data Factory v2 Azure Analysis Services Power BI Database ETL/ELT OLAP Reporting
  • 16.
    Phase 1: Lift& Shift Starting point SQL 2008 R2 Integration Services Analysis Services Reporting Services Phase 1 SQL Server IaaS Integration Services IaaS Analysis Services IaaS Reporting Services IaaS Phase 2 SQL Server IaaS Azure Data Factory v2 with SSIS Azure Analysis Services Power BI Phase 3 Azure SQL DB / DWH Azure Data Factory v2 Azure Analysis Services Power BI Database ETL/ELT OLAP Reporting
  • 17.
    Main objectives ▪ Migratesolution to VM(s) ▪ Take advantage of Azure IaaS services ▪ Setup secure and manageable environment ▪ Gain confidence
  • 18.
    Lift & Shiftoptions ▪ Backup-restore ▪ Data Migration Assistant ▪ IaaS – Windows template ▪ IaaS – SQL Server templates ▪ SQL Database Managed Instances ▪ License considerations
  • 19.
    Data Migration Tools ▪Data Migration Assistant ▪ Supports both scenarios ▪ On-premises -> IaaS -> PaaS ▪ Azure Database Migration Service ▪ Managed service for migrations at scale
  • 20.
    PaaS services coverage Covered ▪Database Engine (SMP / MPP) ▪ Analysis Services ▪ Integration Services (via ADF) Not covered ▪ SQL Server Agent ▪ Reporting Services ▪ Master Data Services ▪ Data Quality Services …Managed Instance?
  • 21.
  • 22.
    Infrastructure recommendations ▪ Planfor PaaS from the beginning ▪ Decouple services to separate VMs ▪ Use Premium managed disks ▪ Less burden with storage accounts (leftovers) ▪ Consider 2/4 TB disks (P40/P50) ▪ Azure Backup supports up to 16 data disks ▪ Azure Backup + Recovery Vault ▪ Use both (VM + disks/data)
  • 23.
    Licensing considerations ▪ Pay-as-you-Govs BYOL ▪ BYOL: 2008 R2+ are supported ▪ Consider SQL Server Developer Edition ▪ Consider memory-optimized VMs (G- / M-) for Analysis Services Tabular ▪ Dev/Test subscriptions
  • 24.
    Cost control recommendations ▪Start small ☺ ▪ Cloudyn – Azure Cost Management ▪ Consider Auto-Shutdown ▪ Reserved instance? Commitment vs flexibility ▪ DEMO – Azure Portal
  • 25.
    Phase 2: Transition Starting point SQL2008 R2 Integration Services Analysis Services Reporting Services Phase 1 SQL Server IaaS Integration Services IaaS Analysis Services IaaS Reporting Services IaaS Phase 2 SQL Server IaaS Azure Data Factory v2 with SSIS Azure Analysis Services Power BI Phase 3 Azure SQL DB / DWH Azure Data Factory v2 Azure Analysis Services Power BI Database ETL/ELT OLAP Reporting
  • 26.
    Main objectives ▪ Startusing PaaS services ▪ Optimize IaaS infrastructure ▪ Understand and control costs
  • 27.
    PaaS recommendations ▪ Startwith Analysis Services – quick win ▪ Consider Azure SQL DB for DWHs < 1 TB ▪ Redesign ETL/ELT in phases ▪ Start with SSIS runtime ▪ Rewrite to pure ELT afterwards ▪ Orchestration instead of linear pipelines
  • 28.
    Analysis Services deploymentoptions Multidimensional ▪ On-premises ▪ IaaS (BYOL) ▪ IaaS (SQL Server template) Tabular ▪ On-premises ▪ IaaS (BYOL) ▪ IaaS (SQL Server template) ▪ Azure Analysis Services ▪ Power BI Premium (de facto)
  • 29.
    Azure Analysis Services ▪Scaling up / out – finally! ▪ Pricing tiers ▪ Replicas ▪ Managed backups ▪ Expensive service – use with care ▪ REST API ▪ Can be automated via Logic Apps / Functions
  • 30.
  • 31.
    Azure Data Factoryv2 ▪ Great orchestration tool with simple GUI ▪ Integrated Git support – easy versioning ▪ Runs SSIS (lots of other engines as well!) ▪ Supports ARM templates ▪ Not an ETL tool actually!
  • 32.
    Power BI vsSSRS ▪ Power BI has limited support for pinning SSRS reports ▪ Paginated reports likely to be supported in future Power BI releases ▪ Plan for leveraging Power BI from moment of IaaS migration ▪ Keep SSRS on small separate VM if needed
  • 33.
    Phase 3: CloudNative Starting point SQL 2008 R2 Integration Services Analysis Services Reporting Services Phase 1 SQL Server IaaS Integration Services IaaS Analysis Services IaaS Reporting Services IaaS Phase 2 SQL Server IaaS Azure Data Factory v2 with SSIS Azure Analysis Services Power BI Phase 3 Azure SQL DB / DWH Azure Data Factory v2 Azure Analysis Services Power BI Database ETL/ELT OLAP Reporting
  • 34.
    Main objectives ▪ Operatepure PaaS solution ▪ Get ready for Azure DWH with ELT ▪ Enable Big Data / real-time scenarios
  • 35.
    Main activities ▪ Migratingto Azure SQL Database ▪ Getting rid of SSIS / ETL ▪ Rewriting new ELT orchestrations ▪ Minimizing VM workloads
  • 36.
    Beyond classical DWH/BI ▪HDInsight ▪ Data Lake Storage / Analytics ▪ Databricks ▪ .. Data Factory to orchestrate them all!
  • 37.
    Wrap-up ▪ Phased approach ▪Organization needs to learn new ways ▪ Not too fast.. ▪ ..but it’s just the right time to start!
  • 38.
  • 39.