5. ANZ Testing
Symposium 2015
While companies are flocking in ever-increasing numbers to cloud’s “everything-as-a-service” promise,
a reliance on traditional, in-house IT systems and applications still exists.
The Legacy Complication…and need for multi-modal IT
6
Channels & Customer Experience
“High Speed Agility”
DECOUPLING
Core Services
“Low Speed Stability”
Partners
Channels & Customer Experience
Core Services
Typical
project
Front
Back
- Today -
Cost & Pace of change led by front & backend changes
- Tomorrow -
Pace of change led by front end changes only; backend work is separate.
• High flexibility to customer needs
• Short-time to market
• Differentiation from competition
• Back-end for internal process
• High level of standardization
• Efficiency and low cost
6. ANZ Testing
Symposium 2015
Difference between Agile and DevOps
6
Wall of
Conflict
Wanting
Change
Development
• Create effective change
• Add/Modify features
Wanting
Stability
Operations
• Create stability
• Enhance services
Wall of
Conflict
Wanting
Flexibility
Customers
• Create flexibility
• Improve time to market
Agile
Development
DevOps
Benefits:
Alignment between Business & IT
Flexibility
More effective solutions
Reduction in Risk
(Speed to market – through smaller
batches)
Benefits:
Speed to Market
Increased Throughput
Reduction in Risk
Faster Feedback
(Reduced Cost)
8. ANZ Testing
Symposium 2015
Benefits Dependency Network for Continuous Delivery
Vision Objectives Benefits
Implement a
DevOps enabled
optimised IT
delivery
organisation
Increase Business
Agility
Reduced requirement creep (largely a function of development schedule)
Reduced cost of experimentation - in turn reducing the investment hurdle challenge (not rate)
Reduce Time to
Market
Increased velocity and faster time to market
Reduced risk of building the wrong thing (and not finding out until too late)
Reduced WIP
Reduce Running
Costs
Reduced deployment effort (cost)/Reduced deployment duration (time)
Reduced risk of deployment failureIncrease Delivery
Productivity
Reduced idea - realisation cycle time
Continuous Improvement of delivery (as we cycle many times, rather than once a year)
Productivity (as a function of specialisation/standardisation)
Earlier regression test through CI/shift left
Faster defect resolution (function of smaller batch sizes and faster feedback cycle -> less
triage)
Reduced production outages due to badly planned/executed changes through improved ways of working,
automation and continuous deployment, self-healing architectures, modularised IT, consistent infrastructure
across environments
Reduced production downtime to support releases
14. 14
Accenture DevOps Platform
Pre-configured tools, environments, automation blueprints and security features to increase the speed of adoption
Cloud-hosted development
environment integrated to
source code repository via
code review server
Autonomous build and
testing of code submissions
ContinuousIntegration
Quality gates enforced using
standardized tests against re-
usable components and
libraries Continuous Delivery
Managed roll-out of new
features via A/B testing
ContinuousDeployment
Cloud-based hosting and access to application services through Accenture Cloud Platform
Code Analysis
(Sonar and Black
Duck)
Integrated
Development
Environment
(Eclipse, X-code)
Source Code
Repository
(Git / Gerrit)
Continuous
Integration
(Jenkins)
Deploy
(Chef and Docker)
Test
(Selenium, Grid,
Cucumber)
Enabling tools
Development
Source
Control
Code
Review
Build
Code
Quality
Deploy Testing
A/B
Test
Release
https://www.youtube.com/watch?v=cI1RpzdsDQI - https://github.com/accenture
16. ANZ Testing
Symposium 2015
The Delivery models are evolving
Model 1:
Automated Application
Deployment into long-lived
Environments
Model 2:
Cloud provisioning of
Environments is part of the
Deployment process achieving
zero-downtime deployments
Model 3:
Container-based deployment of
Microservice based application
services