Rafal Leszko (@RafalLeszko, rafal.leszko@gmail.com)
Continuous Delivery: Jenkins,
Docker and Spring Boot

Senior Software Engineer

I used to work for

Author

Jez Humble

What is
Continuous Delivery?

Continuous Delivery
"The ability to get changes-features, configuration
changes, bug fixes, experiments - into production or
into the hands of users safely and quickly in a
sustainable way"

Have you heard the
story of
Yahoo and Flickr?

● 11.6 seconds Mean time between deployments (weekday)
● 1,079 Max # of deployments in a single hour
● 10,000 Mean # of hosts simultaneously receiving a
deployment
● 30,000 Max # of hosts simultaneously receiving a
deployment

So, how does the
"traditional" software
delivery look like?

"Traditional" Delivery
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

How can we
automate the
process?

Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

What is CI?

Continuous Integration

https://github.com/Dynatrace/ufo

Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

"last mile"
Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

Automated Tests
"last mile"
Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

Automated
Testing

"Acceptance testing is a test performed to determine if
the business requirements or contracts are met. It
involves black-box testing against the complete system
from the user perspective and its positive result should
imply the acceptance for the software delivery."

Should acceptance
testing be
automated?

If acceptance testing
is automated, what is
the role of a QA?

Automated Tests
"last mile"
Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

Configuration
Management
+
MonitoringAutomated Tests
"last mile"
Continuous
Integration
Customer
"Agile" team
Centralized QA IT Operations
Analysis + Design
Development
Testing + Showcase
Integration + QA
Release and operation
Acceptance Tests

Continuous Delivery Pipeline

Checkout
Compile
Unit Tests
Deploy to Staging
Acceptance Tests
Release
Smoke Test

What is Spring Boot?

"Spring Boot makes it easy to create stand-alone,
production-grade Spring based Applications that you
can 'just run'. We take an opinionated view of the
Spring platform and third-party libraries so you can get
started with minimum fuss. Most Spring Boot
applications need very little Spring configuration."

Spring Boot features
● Create stand-alone Spring applications
● Embed Tomcat or Jetty directly (no need to deploy WAR files)
● Provides a 'starter' POM to simplify the Maven configuration
● Automatically configure Spring whenever possible (CoC)
● Provide production-ready features such as metrics, health
checks and externalized configuration
● No code generation and no XML configuration

dev machine
github server
jenkins server
push
pull

Checkout
Compile
Unit Tests
Deploy to Staging
Acceptance Tests
Release
Smoke Test

issues?
● Each server has to be configured
● Different environment in staging and production
● Different Java version for different applications
● Applications using the same ports
● Versioning

What is Docker?

server
application
app.jar

server
application
app.jar
java -jar app.jar

server
application
app2.jar
java -jar app.jar
application
app.jar

treat your servers like cattle,
not pets

Issues with virtualization
● Low performance: Virtual machine emulates the whole
computer architecture, so there is a significant overhead
● High resource consumption: Emulation requires a lot of
resources and has to be done separately for each application.
● Large image size: Each application is delivered with the full
operating system

Dockerfile:
FROM frolvlad/alpine­oraclejdk8:slim
ADD build/libs/vox­0.0.1­SNAPSHOT.jar app.jar
ENTRYPOINT ["java","­jar","app.jar"]

FROM frolvlad/alpine­oraclejdk8:slim
ADD build/libs/vox­0.0.1­SNAPSHOT.jar app.jar
ENTRYPOINT ["java","­jar","app.jar"]
docker build ­t <image_tag> .
Dockerfile:

$ docker run ­­name <container_name> <image_tag>

$ docker run ­­name <container_name> <image_tag>

Docker registry
● registry for docker
images

What is an artifact
repository?

How to push Docker
image to the registry?

image_tag = <registry_address>/<image_name>:<version>

image_tag = <registry_address>/<image_name>:<version>
docker push <image_tag>

image_tag = <registry_address>/<image_name>:<version>
docker push <image_tag>
docker pull <image_tag>

Port Mapping
$ docker run ­p 8080:8080 ­­name <container_name>
<image_tag>

External Docker Host
$ docker run ­p 8080:8080 ­­name <container_name>
<image_tag>
$ docker run ­H 192.168.0.15:2375 ­p 8080:8080
­­name <container_name> <image_tag>

Example
$ docker run ­H 192.168.0.15:2375 ­p 8080:8080 ­d
­­name exchange­rate 192.168.0.20/exchange­rate:15

dev machine
github server
jenkins server
push
pull
staging machine
production machine
deploy
acceptance test
deploy
smoke test
docker registry
push
pull
pull

Manual Testing

Legacy Projects

Performance Tests

Is Continuous
Delivery just a tool?

"continuous"

Trunk-based development

Features branches

Feature Toggle

Easy?

Questions?