The Northwestern Mutual Life Insurance Company – Milwaukee, WI A MongoDB Journey..... Kurt Schwanz Liz Kegel 10.28.2019
Itinerary WHO IS NORTHWESTERN MUTUAL AND CONTENT SERVICES APPLICATION OVERVIEW OUR JOURNEY TO ATLAS INSIGHTS Q&A 2
3 Show of Hands
Who is Northwestern Mutual? 4
Revenue $28.5 billion #104 on FORTUNE 500 9,000+ Financial Professionals 5,900+ Employees Headquartered in Milwaukee, Wisconsin Begun in 1857, Northwestern Mutual is a Fortune 500 company with more than 4.5M clients. Figures as of Dec. 31, 2018, unless otherwise noted
Investing In Innovation Northwestern Mutual Data Science Institute - a $40 million partnership to attract and develop talent and advance the power of data
Who Are We? 7
• Support multiple tools and content repositories. – 500M+ Documents under management – 50+ Business Areas – 1M+ users when we include our field representatives and clients • Follow a DevOps practice where we support the full stack of our applications. 8 Content Services TO BE THE “GO-TO” SOURCE FOR HOW THE COMPANY INTERACTS WITH ITS CONTENT THE DIGITAL WORKPLACE - CONTENT SERVICES
The Journey Begins “The journey of a thousand miles starts with a single step.” -Lau Tzu 9
10 A Quest for Consolidation Pain points: • Performance • Silos / User Experience • Scalability
11 What is Access Content Anywhere (ACA) Central Hub Secure & Compliant Content Life Cycle Collaboration
• Build or Buy? • Researched solutions led to the selection of Nuxeo 3rd party framework and choice of MongoDB. – Schema flexibility – Scalability • Direction to move to the cloud aided by digital transformation. 12 The Fork in the Road..
13 Access Content Anywhere – Conceptual Overview Applications Cloud Object Business Document Stores On Premise Business Document Stores Third Party Business Document Stores Access Content Anywhere Platform Search Retrieve Add Update
14 The Landscape Technology Transformation and ”Re-tooling” brought new ways of working and challenges to the team. • Navigating digital transformation across the organization. • IaaS pattern • Deploying infrastructure as code was new to the team and company. • Needed to learn, deploy, and become proficient with the new ACA Platform.
15 The Journey Ahead Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
16 And We Are Off! Destination 1 Self -Managed MongoDB
17 Leg 1 of the Journey – MongoDB IaaS Deployment ACA Platform MongoDB Supporting Components Content Application Platform Ops Manager Number of VMs 4 OpsMgr x 2 env = 8 TOTAL support hrs per quarter = 100 TOTAL = 20 TestDev Stage ProdNon-Prod Prod 3 MongoDB x 4 env = 12 + 5 hours per VM x
• Allow time for organizational change. • Self-Managed MongoDB requires tuning and you need to have the right resources. • Implement SSL from the start. Retro-fitting is trickier. • Validating your backup and restore process and failover is crucial. • Leverage MongoDB Professional Services. 18 Leg 1 – Self Managed MongoDB - Insights
19 The Journey Continues Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
Enterprise Data Services 20 Leg 2 of the Journey – MongoDB OpsMgr Data as a Service Transition ACA Platform – Content Services MongoDB Supporting Components Content Application Platform Ops Manager Number of VMs Dev Test Stage Prod TOTAL = 12 3 MongoDB x 4 env = 12 0 OpsMgr x 2 env = 0 + 5 hours per VM x TOTAL support hrs. per quarter = 60
• If you have a centralized data services function leverage it. • It can take time for organizations to accept NoSQL databases and be willing to bring under a shared services model. • Moving from one OpsMgr to another is relatively simple. 21 Leg 2 - Interim State - Insights
22 Are We There Yet? Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
ATLAS....T The Fellowship of the Ring 23
• Decision coincided with an upgrade of the 3rd party Content Services framework requiring a newer version of MongoDB. • SaaS solutions were becoming accepted. • DevOps team could focus on new development instead of support. • Support for multiple Availability Zones and improved backup and restore capabilities. • MongoDB had earned our trust. 24 Why Atlas
z ATLAS 25 Leg 3 of the Journey – Fully Implemented Atlas ACA Platform – Content Services MongoDB Supporting Components Content Application Platform TOTAL support hours per quarter = 0 Number of VMs 0
26 How Did We Get There? MongoMirror was used to migrate our existing data to Atlas • Set up options were straightforward • Fear and trepidation were unfounded • Very pleased with the speed of the migration – 63 Million documents, 81 GB database took 3.5 hours • Thanks to oplog-tail feature, no downtime was experienced. Performance Testing • Response times were the same (and better for some tests) than our IaaS infrastructure.
Authentication • LDAP was not an option for our enterprise • Needed to work through and document acceptable compensating controls. • Waiting for MongoDB to release Federated Identity Management and SSO! Internal Constraints • Security reviews and sign-off took longer than anticipated. 27 Speed Bumps Are Unavoidable
Understand your enterprise security posture before jumping in. • Know your data classifications and applicable regulations • Understand your firewall considerations and permissions • Authentication Mongo-Mirror is your friend for data migration! Performance Testing is a given. Leverage MongoDB Support with questions. 28 Leg 3 – MongoDB Atlas Insights
29 You Have Arrived at Your Final Destination Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
As One Journey Ends... A New One Begins The Fellowship of the Ring 30
• Is your organization open to managed database offerings? • Do you have the skill sets and capacity in house to self-support MongoDB? • Do you have a need for multiple Availability Zone coverage for your data and backups? • Organizational change takes time. • Mongo Mirror is your friend. • Leverage MongoDB Support 31 Souvenirs
Questions? 32

MongoDB .local Chicago 2019: A MongoDB Journey: Moving from a relational database to MongoDB Atlas

  • 1.
    The Northwestern MutualLife Insurance Company – Milwaukee, WI A MongoDB Journey..... Kurt Schwanz Liz Kegel 10.28.2019
  • 2.
    Itinerary WHO IS NORTHWESTERNMUTUAL AND CONTENT SERVICES APPLICATION OVERVIEW OUR JOURNEY TO ATLAS INSIGHTS Q&A 2
  • 3.
  • 4.
  • 5.
    Revenue $28.5 billion #104on FORTUNE 500 9,000+ Financial Professionals 5,900+ Employees Headquartered in Milwaukee, Wisconsin Begun in 1857, Northwestern Mutual is a Fortune 500 company with more than 4.5M clients. Figures as of Dec. 31, 2018, unless otherwise noted
  • 6.
    Investing In Innovation NorthwesternMutual Data Science Institute - a $40 million partnership to attract and develop talent and advance the power of data
  • 7.
  • 8.
    • Support multipletools and content repositories. – 500M+ Documents under management – 50+ Business Areas – 1M+ users when we include our field representatives and clients • Follow a DevOps practice where we support the full stack of our applications. 8 Content Services TO BE THE “GO-TO” SOURCE FOR HOW THE COMPANY INTERACTS WITH ITS CONTENT THE DIGITAL WORKPLACE - CONTENT SERVICES
  • 9.
    The Journey Begins “Thejourney of a thousand miles starts with a single step.” -Lau Tzu 9
  • 10.
    10 A Quest forConsolidation Pain points: • Performance • Silos / User Experience • Scalability
  • 11.
    11 What is AccessContent Anywhere (ACA) Central Hub Secure & Compliant Content Life Cycle Collaboration
  • 12.
    • Build orBuy? • Researched solutions led to the selection of Nuxeo 3rd party framework and choice of MongoDB. – Schema flexibility – Scalability • Direction to move to the cloud aided by digital transformation. 12 The Fork in the Road..
  • 13.
    13 Access Content Anywhere– Conceptual Overview Applications Cloud Object Business Document Stores On Premise Business Document Stores Third Party Business Document Stores Access Content Anywhere Platform Search Retrieve Add Update
  • 14.
    14 The Landscape Technology Transformationand ”Re-tooling” brought new ways of working and challenges to the team. • Navigating digital transformation across the organization. • IaaS pattern • Deploying infrastructure as code was new to the team and company. • Needed to learn, deploy, and become proficient with the new ACA Platform.
  • 15.
    15 The Journey Ahead Destination1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
  • 16.
    16 And We AreOff! Destination 1 Self -Managed MongoDB
  • 17.
    17 Leg 1 ofthe Journey – MongoDB IaaS Deployment ACA Platform MongoDB Supporting Components Content Application Platform Ops Manager Number of VMs 4 OpsMgr x 2 env = 8 TOTAL support hrs per quarter = 100 TOTAL = 20 TestDev Stage ProdNon-Prod Prod 3 MongoDB x 4 env = 12 + 5 hours per VM x
  • 18.
    • Allow timefor organizational change. • Self-Managed MongoDB requires tuning and you need to have the right resources. • Implement SSL from the start. Retro-fitting is trickier. • Validating your backup and restore process and failover is crucial. • Leverage MongoDB Professional Services. 18 Leg 1 – Self Managed MongoDB - Insights
  • 19.
    19 The Journey Continues Destination1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
  • 20.
    Enterprise Data Services 20 Leg 2of the Journey – MongoDB OpsMgr Data as a Service Transition ACA Platform – Content Services MongoDB Supporting Components Content Application Platform Ops Manager Number of VMs Dev Test Stage Prod TOTAL = 12 3 MongoDB x 4 env = 12 0 OpsMgr x 2 env = 0 + 5 hours per VM x TOTAL support hrs. per quarter = 60
  • 21.
    • If youhave a centralized data services function leverage it. • It can take time for organizations to accept NoSQL databases and be willing to bring under a shared services model. • Moving from one OpsMgr to another is relatively simple. 21 Leg 2 - Interim State - Insights
  • 22.
    22 Are We ThereYet? Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
  • 23.
  • 24.
    • Decision coincidedwith an upgrade of the 3rd party Content Services framework requiring a newer version of MongoDB. • SaaS solutions were becoming accepted. • DevOps team could focus on new development instead of support. • Support for multiple Availability Zones and improved backup and restore capabilities. • MongoDB had earned our trust. 24 Why Atlas
  • 25.
    z ATLAS 25 Leg 3 ofthe Journey – Fully Implemented Atlas ACA Platform – Content Services MongoDB Supporting Components Content Application Platform TOTAL support hours per quarter = 0 Number of VMs 0
  • 26.
    26 How Did WeGet There? MongoMirror was used to migrate our existing data to Atlas • Set up options were straightforward • Fear and trepidation were unfounded • Very pleased with the speed of the migration – 63 Million documents, 81 GB database took 3.5 hours • Thanks to oplog-tail feature, no downtime was experienced. Performance Testing • Response times were the same (and better for some tests) than our IaaS infrastructure.
  • 27.
    Authentication • LDAP wasnot an option for our enterprise • Needed to work through and document acceptable compensating controls. • Waiting for MongoDB to release Federated Identity Management and SSO! Internal Constraints • Security reviews and sign-off took longer than anticipated. 27 Speed Bumps Are Unavoidable
  • 28.
    Understand your enterprisesecurity posture before jumping in. • Know your data classifications and applicable regulations • Understand your firewall considerations and permissions • Authentication Mongo-Mirror is your friend for data migration! Performance Testing is a given. Leverage MongoDB Support with questions. 28 Leg 3 – MongoDB Atlas Insights
  • 29.
    29 You Have Arrivedat Your Final Destination Destination 1 Self-Managed MongoDB Destination 2 Interim Final Destination Atlas
  • 30.
    As One JourneyEnds... A New One Begins The Fellowship of the Ring 30
  • 31.
    • Is yourorganization open to managed database offerings? • Do you have the skill sets and capacity in house to self-support MongoDB? • Do you have a need for multiple Availability Zone coverage for your data and backups? • Organizational change takes time. • Mongo Mirror is your friend. • Leverage MongoDB Support 31 Souvenirs
  • 32.