EBSCO Digital Transformation with AWS
- 2. ● Largest business unit in EBSCO Industries - over
3,200 employees
● The most highly-used online research platform to
libraries worldwide
● Largest subscription service agency in the world
● Over 150,000 library customers worldwide
● Relationships with more than 95,000 publishers
internationally
- 3. Product Lines
Approaching 400 research
and educational databases
via EBSCOhost with over
2.6 billion records, and 100
million searches per day
- 4. About Kenzan
Core offerings
Application Development and
Consulting, Platform as a Service
(PaaS), and Digital Transformation
Primary Clients
Multi billion dollar companies and
media/content providers such as
Thomson Reuters, Charter &
Cablevision
Locations
Providence (RI), New York (NY),
Denver (CO), and Los Angeles (CA)
Founded in 2004.
DEVOPS
LEADERSHIP
APPLICATION DEVELOPMENT
Architecture, front and back end
development, business analysis
and DevTest.
Platform builds, continuous
delivery and scalable resourcing.
CLOUD
VIRTUALIZATIONExperts and enablers in AWS,
Netflix stack, enterprise
architecture and beyond.
MEDIA INDUSTRY VETS
Migrations, enterprise wide
solutions, digital experts and
thought leaders.
EMPLOYEE-FOCUSED
Collaboration and culture are
key.
FULL SERVICE
CONSULTING
DIGITAL TRANSFORMATION
Developing technology as a core
competency & organizing teams
around business value
- 5. Cloud Transformation =
Culture Transformation
EBSCO Goals
● Customer orientation
● High velocity value creation
● Sustainable acceleration
● Margin Protection
● Transformational software
Aligned Technology, Process
and Organization
Culture
Technology
OrganizationProcess
Cloud and Microservices
Agile
Methods
Inverse Conway’s
Law
An all-in executive lead approach
is required.
- 6. Technology
Organization
Agile
Current Future
● Large deployments
● Tightly coupled
● Monolithic
● Trend to Serverless
● Loosely Coupled
● Independent Domains
● Frequent handoffs CI/CD
● Shared ownership
● T-shaped
● Large non-coding budget
● CI/CD/CO to No DevOPs
● Component Ownership
● Domain expertise
● Phase out of non-coding teams
Aligned to business
High Performance
● Art-centric
● Feature Backlogs
● 2-week iterations
● Monthly/Quarterly Delivery
● Ceremony decrease
● Capability Backlogs
● 1-weeks iterations
● Daily Delivery
Value delivery focused
- 7. Technology Design Center
Evolved and
Differentiated
• Contextual
• Analytic
• Closed-Loop
Fast
• Composable
• NIH/3P
• Independent services
• Container-based
• Leveraged platform
Assume Failure
• Circuit breakers
• Contract tests
• Chaos Engineering
• Blue/Green and Canary
Microservices
Model Characteristics
Technology Decision Tree
Capability available through
third party:
1. Netflix OSS
2. AWS Native Tools and
Services
3. 3rd party Managed
(deployed in AWS)
4. Other Open Source
5. Cloud Native COTS
6. EBSCO built
Capability differentiated and not
available through 3rd party
1. Single consumer: Product
ART build
2. Multi consumer: Platform
ART build
Core Framework Technology Guidance
Limit Tech Sprawl – Distributed Decision Making – Team Independence
Domain Model
Bounded Context
- 9. The Answer:
• Developers build, deploy and
run their software
• Operations automates the
operation, not operate the
operation
• Stop, minimize, outsource or
automate ALL non-coding
functions
The Development Centric Operating Model
The Vision
Maximize the
development of
differentiated software
The Opportunity
Provide unfettered access
to cloud computing
resources and remove all
things that block engineers
and getting software to
production
- 10. ”Bad behavior arises when you abstract
people away from the consequences of
their actions…there is no such thing as
‘DevOps Team!”
- Jez Humble
- 11. Business Alignment
with AWS
• Orientation to revenue and value
• Visible customer touch
• Multivariate tests
• Zero defects
• Continuous operations
• Infrastructure as code
• Minimize non-dev technology
positions
• Retrain/re-assign traditional ops
and QA
• Cloud Ops automates the operation
• Rapid provisioning, access and
change management
GOALS
• 25% increase in new feature dev
• 80% decrease in non-dev labor
• 20 minute goal code complete to live
DYNAMIC HEALTH ACTUALS
• 2 weeks: working prototype
• 9 months: V1.0
• 3 months: V2.0
- 13. All-In Transformation
Timeline of delivery
Hardening
Platform Dev. +
Training
Materials
Validation
Review
architecture
Jan 2017
Mid-Dec 2016
Training
Onboarding
SWAT teams
Jan 2017
Oct 2016
Proposal
2017+Apr 2017
Support
Player /
Coach
Sustain/expand
Transition +
Additional
Projects
May 2017
Create Quick Sustain / Expand
✓
- 16. DevTest BDD
Automated
Testing
Code
Review
GitHub
Flow
Fail-
Forward
Loosely coupled 12-factor applications
Continuous Learning
Logging Metrics Auto
Scaling
APM Dashboard Application
Support
Continuous Operations
Immutable
Infrastructure
Immutable
Artifact
Automated
Testing
Config
Management
Canary/
BlueGreen
A/B
Testing
Continuous Deployment
Continuous Integration
Agile, Lean Culture (Strong Organizational Health)
- 18. Platform, Dev Kit, Training Materials
Create a vision >> Hardening
Platform
Automation
● Github
● Jenkins / JFrog
● ECS / Docker
● Cloud Formation
● SumoLogic/ DataDog
Continuous Delivery
● Jenkins pipeline library
● Immutable Infrastructure
Architecture
● Naming Conventions
● Service Discovery
Platform Dev Kit Training materials
Materials
● Course Catalog
● Onboarding Guides
● Activities & Labs
Projects
● Hello-World
● Reference Architecture
Libraries
● Starters
● Logging
● Metrics
● Circuit Breaking
● Configuration
- 20. Quick Win
Communicate Vision Enlist team Remove Blockers
20
Training
● Teams (3w -> 1.5w)
● Onboarding Guides
● Design-to-deploy
Progress Reports
● Rally Metrics
● Completed/Goals
● Issues/Risks
● Key Decisions
Projects
● Hello World
● Reference Architecture
Course Catalog
● PLATFORM, SERVICE,
UI
● 100, 200, 300 Level
Architecture
● Domain Mapping
● System Diagram
Support Styles
● Player / coach
● Train and assess
● Do and demo
● Try and review
Platform Development
● Deploy to INT, Live
● Canary
Cross team communication
● Weekly emails, DSU
● Chatroom support
- 23. Transformation Timeline
2016 2017 2018
July: Initial technology strategy
July: AWS Seattle briefing
August: Working Netflix
prototype
October: Validated reference
arch
November: Kenzan
introductions
December: Budget to build Dyn
Health
January: Kenzan led initial training
February: Working Dyn Health
prototype
August: Executive “All In”
agreement
August: Start of Researcher
prototype
September: CO Simulation
October: DH Version 1.0
November: New org/ART structure
January: Increase R&D spend to
50% on AWS
February: Researcher “preview”
Q2: Initial API-based public
services on AWS
Q3: DynamedPlus on AWS
Q4: 75% on AWS, all content on
AWS
- 24. Recommended Practices
Find experienced partners
• AWS: tech releases, well
architected
• Kenzan: Microservices
expertise, Netflix OSS
experience
Avoid Porting
Operating costs and
extraneous code
$
Velocity
Frameworks and
guidance create
velocity
Feedback
Customer feedback/Live
code on Day 1
You aren’t going to need it
Platform and services follow
YAGNI principle
Baseline model and
rapid evolution
Organizational,
budgetary and
technology
Wins
Working software
wins
- 25. Key Learnings
• No perfect time to start
• Firewall AWS work from on premise work
• Tech transition from on-premise .NET a none issue
• Teams left behind
• Teams like the Continuous Operations Model
• Start refactoring on day 1
• AWS Teams are innovating more
• Cost management, telemetry and automation on
day 1