Continuous Deployment to the cloud

Continuous Deployment
of your Application
By Cora Iberkleid
@ciberkleid
1
By Marcin Grzejszczak
@mgrzejszczak

Senior Platform Architect at Pivotal
Working on
● Pivotal Cloud Foundry
● Cloud Native Architecture
● Dev Practices Modernization
Twitter: @ciberkleid
About us

Spring Cloud developer at Pivotal
Working on
● Spring Cloud Sleuth
● Spring Cloud Contract
● Spring Cloud Pipelines
About us
Twitter: @mgrzejszczak
Blog: http://toomuchcoding.com

What problem are we trying to solve?

The necessity to create a deployment
pipeline from scratch for each new project

Spring Cloud Pipelines
• opinionated template of a deployment pipeline
• based on good practices from different projects
• we believe in the “Infrastructure as Code” approach
• in case of automation server going down you can recover everything in no time
• the automation server setup should be automated too!
• you can even write tests for your pipelines

• we support following automation servers out of the box
• Concourse
• Jenkins using Jenkins Job DSL plugin
• Jenkins using Jenkinsfile with Blue Ocean
• logic of all steps done via Bash scripts
• you can convert the internals to suit your needs
• you can use whatever automation server you want
• supports Maven & Gradle
• Over 150 Bash tests (using https://github.com/sstephenson/bats)

Spring Cloud Pipelines - PaaS Types

Spring Cloud Pipelines - CF example

Spring Cloud Pipelines - Concourse

Spring Cloud Pipelines - Jenkins Job DSL

Spring Cloud Pipelines - Jenkinsfile & Blue Ocean

WHAT ARE WE GOING TO WORK WITH?

Demo - flow
GITHUB-WEBHOOK GITHUB-ANALYTICS
AMQPHOOK

Why do we deploy software?
• We’re paid for delivering business value
• Features are done when they are deployed to production
• It’s better to deploy frequently for faster feedback
• It’s better to fail the deployment pipeline as fast as possible
• Deployments should take place in a standardised, automated fashion
• Your deployment pipeline should test rollback
• That way you can perform A/B or zero downtime deployment to production

• We’re paid for delivering business value
• Features are done when they are deployed to production
• It’s better to deploy frequently for faster feedback
• It’s better to fail the deployment pipeline as fast as possible
• Deployments should take place in a standardised, automated fashion
• Your deployment pipeline should test rollback
• That way you can perform A/B or zero downtime deployment to production

Problem - slow feedback
• Nobody wants to wait until the end of the pipeline to see that something is not
working fine
• We want to fail fast - even during build time
• If integration is faulty
• If our API is backward incompatible
• There should be no need to wait until end to end tests are executed

Solution - fail fast
• We’re running unit and integration tests during build time
• To test faulty integration we use Spring Cloud Contract for HTTP / messaging
• Producer defines a contract
• From the contract
o tests are generated to see if the producer is not lying
o stubs are generated for consumers to use
• Consumer can reuse it without running the producer
• Fail at build time if integration fails (fail fast!)
• All stubs reside in Nexus / Artifactory (thus can be reused)

CONTRACT TESTS DEMO
(BREAKING CONSUMER)

Initial contract (producer)
Body contains username and
repository

Generated test (producer)
Body contains username and
repository

Passing contract test (consumer)

Breaking contract (producer)
Was username and repository
and we’ve made a breaking
change by converting those to
user and repo

Installing broken stubs locally (producer)
New stubs with backward
incompatible changes
installed in Maven local

Broken contract test (consumer)

Broken contract test(consumer)
Expected repository and
username but got repo
and user

CONTRACT TESTS DEMO
(BREAKING PRODUCER)

Demo - backward incompatible API change
GITHUB-ANALYTICS V1
with DELETE
@ /issues
Contracts
V1
with
DELETE @
/issues
Generated test
sends DELETE
to /issues

Contract for deletion of issues

Generated test for deleting issues

GITHUB-ANALYTICS V1
with DELETE
@ /issues
Contracts
V1
with
DELETE @
/issues
Generated test
sends DELETE
to /issues
GITHUB-ANALYTICS V2
removed DELETE
@ /issues
Contracts
V2
removed
DELETE @
/issues
No generated
test for this
endpoint

Backward incompatible changes of the API
We delete
the
contract
and the
DELETE
endpoint

GITHUB-ANALYTICS V1
with DELETE
@ /issues
Contracts
V1
with
DELETE @
/issues
Generated test
sends DELETE
to /issues
GITHUB-ANALYTICS V2
removed DELETE
@ /issues
Contracts
V2
removed
DELETE @
/issues
No generated
test for this
endpoint
GITHUB-ANALYTICS V2
removed DELETE
@ /issues
Contracts
V1
with
DELETE @
/issues
Generated test
sends DELETE
to /issues

Broken build Current implementation (V2)
does not support old
contracts (V1)

Broken - generated test from old contracts (V1)

STANDARDISATION GIVES LOWER
SUPPORT COSTS AND MORE CHANCES
OF PEOPLE ROTATION

Solution - PaaS & tooling
• Use a PaaS to standardise the way you deploy and monitor your software
• Spring Cloud Pipelines uses Cloud Foundry [1] or Kubernetes [2] as a PaaS
implementation
• For the demo purposes we’re using PCF Dev [3] or Minikube [4]
• PaaS abstracts the application governance from underlying infrastructure
• you can deploy, scale, manage applications in the same way if PaaS was running on your laptop,
Amazon, Google, Azure etc.
• Database schema upgrade is done via tools like Flyway [5] or Liquibase [6]
[1] https://www.cloudfoundry.org, [2] https://kubernetes.io/
[3] https://pivotal.io/pcf-dev, [4] https://github.com/kubernetes/minikube
[5] https://flywaydb.org/, [6] http://www.liquibase.org/

After the application got deployed to test environment
• The database schema gets updated upon application startup
• We run a handful of smoke tests to see if crucial parts of our application are
working fine
• We want to test if the app is properly packaged
• The application is surrounded by stubs - no real integrations take place
• Spring Cloud Contract Stub Runner Boot app is responsible for serving stubs

Problem - rollback DB
• Deployment pipelines should test whether the application can be rolled back
• Rolling back database changes is extremely difficult
• Especially if you deploy once every 6 months (the number of changes is gigantic)
• How can you roll back a deletion of a column?

Solution - application rollback
• The easiest solution is… NOT TO DB ROLLBACK
• Perform only backward compatible changes (always add data)
• Or perform backward incompatible changes in a backward compatible way [1]
• Roll back the application only (the JAR)
• The application and DB changes need to be backward compatible
• That way you can ensure that two applications (old / new versions) can run
simultaneously at the same time
[1] https://spring.io/blog/2016/05/31/zero-downtime-deployment-with-a-database

BACKWARD INCOMPATIBLE
DB CHANGE DEMO

Demo - backward incompatible DB change
GITHUB-ANALYTICS V1
CODE
with repository
DB
V1
with
repository

Demo - initial DB state (Flyway)

GITHUB-ANALYTICS V1
CODE
with repository
DB
V1
with
repository
GITHUB-ANALYTICS V2
CODE
with repo
DB
V2
with
repo

GITHUB-ANALYTICS V1
CODE
with repository
DB
V1
with
repository
GITHUB-ANALYTICS V2
CODE
with repo
DB
V2
with
repo
GITHUB-ANALYTICS V1
CODE
with repository
DB
V2
with
repo

New application (V2) works
against new database (V2)

Old application (V1) can’t
work with new database (V2)
Deploy (V1) and connect to
the new database (V2)

Demo - errors in the app logs Old version can’t insert
data to a missing DB
column

Continuous Deployment to the cloud

Problem - end to end tests
• End to end tests are slow and brittle
• QA department writes an E2E for every feature we have
• E2E environment setup
• one environment shared between all applications?
• one environment per application?
• Surrounding apps should be deployed in
• production versions?
• development versions?

Solution - don’t do E2E?
• Regardless of the time spent on QA / UAT you can still have bugs on
production
• Assuming that you ...
• embrace failure
• introduce monitoring of business key performance indicators (KPIs)
• introduce alerting over the metrics
• ensure that you can rollback on production
• … you could stop doing any end to end tests

• Deploy to stage and running e2e tests are manual steps
• you have to wait for your turn for the env
• some manual work has to be done to purge stale data etc.

Problem - deployment to production
• We don’t want to deploy the application to production at night
• We want to treat a production deployment like any other deployment
• We’d like to be able to perform A/B testing and zero downtime deployment
• We’d like to easily rollback when sth goes wrong

Solution - PaaS + SC Pipelines
• Our application has KPI monitoring in place
• Alerting are set for KPIs
• It has been tested that the application can be easily rolled back
• PaaS (CF or K8S) can take care of zero downtime deployment

• prod-deploy deploys the pipeline version of the app to production next to
the current production version
• Once deployed we tag the repo with prod/VERSION_NUMBER
• prod-complete allows to stop the old instance and leave only the new one
running
• prod-rollback deletes the new instance, removes the
prod/VERSION_NUMBER tag and leaves only the old one running

Two versions running at the same time
Two versions registered
under same hostname
Old instance
New instance

MIGRATING TO SPRING CLOUD
PIPELINES DEMO

Application Refactoring for
A Story in Three Acts

Unless otherwise indicated, these slides are © 2013-2017 Pivotal Software, Inc. and licensed under a Creative Commons Attribution-
NonCommercial license: http://creativecommons.org/licenses/by-nc/3.0/
Main Characters
84
Greeting UI Fortune
Service
Fortune DB
Simon

Supporting Characters
85
The SCS Triplets
and their pet Rabbit
Greeting UI Fortune
Service
Fortune DB
Simon

The Setting

Story Line
87
Greeting UI
Simon

Story Line
88
Greeting UI
Simon
Hey Greeting!

Story Line
89
Greeting UI
The SCS Triplets
Simon
Hey Greeting!

Story Line
90
Greeting UI
The SCS Triplets
???
Simon
Hey Greeting!

Story Line
91
Greeting UI
The SCS Triplets
!!!
Fortune
ServiceSimon
Hey Greeting!

Story Line
92
Greeting UI
The SCS Triplets
Fortune
ServiceSimon
Hey Greeting!

Story Line
93
Greeting UI
The SCS Triplets
Fortune
Service
Fortune DB
Simon
Hey Greeting!

Story Line
94
Greeting UI
The SCS Triplets
Fortune
Service
Fortune DB
Simon
Hey Greeting!

Story Line
95
Greeting UI
The SCS Triplets
Fortune
Service
Fortune DB
Simon
Hey Greeting!

Story Line
96
Greeting UI
“Do What Works”
Simon

Story Line
97
Greeting UI
Simon
Great advice!
Who said it?

Story Line
98
Greeting UI Fortune
ServiceSimon
Great advice!
Who said it?

Story Line
99
Greeting UI Fortune
ServiceSimon
Great advice!
Who said it?

Story Line
100
Greeting UI Fortune
Service
Fortune DB
Simon
Great advice!
Who said it?

Story Line
101
Greeting UI Fortune
Service
Fortune DB
Simon
Great advice!
Who said it?

Story Line
102
Greeting UI Fortune
Service
Fortune DB
Simon
It’s not that
hard a question...

Story Line
103
Greeting UI Fortune
Service
Fortune DB
Simon
It’s not that
hard a question...
The SCS Triplets

Story Line
104
Greeting UI Fortune
Service
Fortune DB
Simon
I just want to
know...
The SCS Triplets

Story Line
105
Greeting UI Fortune
Service
Fortune DB
Simon
PIPELINE ALL
THE THINGS!!!
The SCS Triplets

Act I: The Scaffold
107
● Objective:
○ Automate the deployment of greeting-ui and fortune-service
a.k.a. Make the pipelines run green
○ Establish environments (PCF spaces) in accordance with SCP
recommendations
● Approach:
○ Minimum changes to apps (scaffolding, not code)
○ Create a pipeline for each
○ Create PCF spaces
○ Run the pipelines

Greeting and Fortune
● Spring Boot apps
● Source code on GitHub
● Maven
● No tests
● Use 5 services
○ MySQL database (for Fortune)
○ Spring Cloud Services: Eureka, Hystrix/Turbine, Config Server
○ Rabbit - for Config Server refresh trigger propagation

● Add 2 new branches to the app repos
App Scaffold
109

● Add a Maven wrapper
○ mvn -N io.takari:maven:wrapper
App Scaffold
110

● Add artifact repo info to app pom.xml (and ~/.m2/settings.xml)
○ distributionManagement.repository.id
○ distributionManagement.repository.url
App Scaffold
111

● Add artifact repo info to app pom.xml
○ distributionManagement.repository.id
○ distributionManagement.repository.url
● git push!
App Scaffold
112

Pipeline Configuration - Git and Mvn
113
● Create a credentials file for each app (copy sample in SCP)
○ Git url and branch of your app
○ Git url and branch of the SCP code base (cf or v1.0.0.M8, coming soon!)
○ Your git name, email, and private key
○ Your maven repo info and credentials

Build - Success!
114
● At this point, the Build jobs should all succeed
● App builds, jar uploads to maven repo, git tagged as “dev/<version>”
● The api compatibility job runs “mvn clean verify -P apicompatibility”, but the
profile does not exist yet, so it really isn’t doing much
● Takes place on Concourse worker

● Create app manifest.yml
● Create a new manifest for SCP (e.g. sc-pipelines.yml)
○ Provide info for SCP to provision services on CF (test* & stage)
● Git push!
App Scaffold
115
test:
services:
- name: fortune-db
type: broker
broker: cleardb
plan: spark
- name: config-server
type: broker
broker: p-config-server
plan: standard
params:
git:
uri: https://github.com/ciberkleid/app-config
useExisting: true
...

Pipeline Configuration - CF
116
● Add general CF info to the credentials file
○ paas-type: cf
○ pipeline-descriptor (the sc-pipelines manifest)
○ paas-hostname-uuid (to avoid route collision)
● Add target & login info for Test, Stage and Prod
○ For Test, you specify a prefix. SCP appends the app name

Cloud Foundry Setup: Test, Stage, Prod
117
● Create the Test, Stage, and Prod spaces
● Manually provision services in Prod
○ And maybe Stage
○ But not Test

Full Pipeline - Success!
118
● App deployed to PCF through Prod
● Git tagged as prod/<version>
● Ability to trigger rollback through Concourse

NonCommercial license: http://creativecommons.org/licenses/by-nc/3.0/ 119

NonCommercial license: http://creativecommons.org/licenses/by-nc/3.0/ 120

Act I: The Scaffold
121
● Objective:
recommendations
● Approach:
○ Minimum changes to apps (scaffolding, not code)
○ Create a pipeline for each
○ Create PCF spaces
○ Run the pipelines

Is That Enough?
122

Nope!
123
Same for apicompatibility
and e2e

Act II: The Test
125
● Objective:
○ Increase effectiveness of and confidence in the pipelines
○ Standardize approach to testing
● Approach:
○ Use mvn profiles to control test executions
○ Add/organize tests in agreement with SCP framework
○ Enable DB backward compatibility testing

Test Control - Maven Profiles
126
● Add profiles to the app pom.xml
○ default
○ apicompatibility
○ smoke
○ e2e
<profile>
<id>default</id>
<activation>
<activeByDefault>true</activeByDefault>
</activation>
<build>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-surefire-plugin</artifactId>
<configuration>
<includes>
<include>**/*Tests.java</include>
<include>**/*Test.java</include>
</includes>
<excludes>
<exclude>**/smoke/**</exclude>
<exclude>**/e2e/**</exclude>
</excludes>
</configuration>
</plugin>
<plugin>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-maven-
plugin</artifactId>
</plugin>
</plugins>
</build>
</profile>

Test Control - Organize Tests
127
● Organize tests into packages matching the profiles
● Add tests for each profile

Test Coverage - Fortune Smoke
128
fortune-service
fortune-db

Test Coverage - Fortune E2E
129
greeting-ui
fortune-db
fortune-service
Assuming same domain...

Test Coverage - Greeting Smoke
130
greeting-ui
circuit breaker fallback

Test Coverage - Greeting E2E
131
greeting-ui
fortune-service

Enable Database Rollback Testing
132
● Flyway
○ OSS DB migration tool
○ For relational databases
○ Convention over configuration
○ Database versioning
○ Migration on app startup (good for CD!)

DB Schema Creation With Flyway
133
● Disable JPA auto-generated schemas; let flyway do it
● Follow flyway conventions for SQL file name
● Tada! Schema is now versioned!

DB Schema Population With Flyway
134
● Add INSERT statements to V1__init.sql
● Data is populated upon app startup, but only once per DB version

Act II: The Test
135
● Objective:
● Approach:
○ Use mvn profiles to control test executions
○ Add/organize tests in agreement with SCP framework
○ Enable DB backward compatibility testing

Is THAT Enough?
136
● Good:
○ You are in a great position to include well-organized, robust testing and
derive a high level of confidence from your pipelines
○ You can ensure safe rollbacks in case of database schema changes
● Better:
○ Implementing a contract-driven approach will have additional benefits to
your general development practice
○ Inter-team communication will be simpler
○ Contract tests will catch breaking API changes in build rather than stage
○ Troubleshooting will be easier
○ Failure and feedback will be faster

Act III: The Contract
138
● Objective:
○ Increase reliability - strive to maximize pipeline benefits
○ Encourage contract-based programming practices
● Approach:
○ Add contracts, stubs, and a stubrunner
○ Enable API backward compatibility testing

The Greeting/Fortune Contract
139
import org.springframework.cloud.contract.spec.Contract
Contract.make {
description("""
should return a fortune string
""")
request {
method GET()
url "/"
}
response {
status 200
body "foo fortune"
}
}

Spring Cloud Contract Basics
140
● Add the Spring Cloud Contract Maven Plugin to your pom.xml (Fortune)
○ Plugin requires a base class to set up the context and stub out the service
that satisfies the API call (e.g. the DB)
public class BaseClass {
@Before
public void setup() {
FortuneService service = BDDMockito.mock(FortuneService.class);
BDDMockito.given(service.getFortune()).willReturn("foo fortune");
RestAssuredMockMvc.standaloneSetup(new FortuneController(service));
}
}

Spring Cloud Contract Basics
141
● Add the Spring Cloud Contract Maven Plugin to your pom.xml (Fortune)
○ Plugin will auto-generate tests and stubs (and a stub.jar)
○ Stub jar is uploaded to the maven repo, along with the app jar

For Back-Compatibility, Use Prod Contracts
142
<profile>
<id>apicompatibility</id>
<plugin>
<configuration>
<contractsRepositoryUrl>${repo.with.binaries}</contractsRepositoryUrl>
<contractDependency>
<groupId>${project.groupId}</groupId>
<artifactId>${project.artifactId}</artifactId>
<classifier>stubs</classifier>
<version>${latest.production.version}</version>
</contractDependency>
...
In this case, we want the contracts plugin to generate tests
based on contracts outside of the project (the ones from the
latest prod jar on our maven repo). Pipeline injects values in red
dynamically

Greeting: Build with Stubrunner
greeting-ui
or manually
configured WireMock
greeting-ui
stubs
without stubrunner with stubrunner
● Greeting starts an in-process stubrunner that automatically configures
WireMock

Greeting: Integration Tests Aligned with Stubs
144
For in-process stubrunner, the
server-side Wiremock port
For in-process stubrunner, the
client-side endpoint config

Greeting: Test with Stubrunner (standalone)
greeting-ui
greeting-ui
stubrunner
without stubrunner with stubrunner
Note: Update the
greeting smoke tests
accordingly

Home Stretch: Standalone Stubrunner Config
● You need a stubrunner app!
○ Clone spring-cloud-samples/cloudfoundry-stub-runner-boot
○ Build it, and upload the jar to your maven repo
● Your stubrunner app needs a manifest!
○ Put it in your app repo (e.g. greeting-ui repo)
● You need to tell SCP to deploy the stubrunner app!
○ Do it through the pipeline descriptor (e.g. sc-pipelines.yml)

Home Stretch: Standalone Stubrunner Config
● Specify the server side (stubrunner) ports
<properties>
<stubrunner.ids>io.pivotal:fortune-service:_latest:stubs:10000</stubrunner.ids>
</properties>
● Disable Eureka lookup for smoke tests
● Provide URLs to non-8080 stubrunner ports explicitly

Act III: The Contract
148
● Objective:
● Approach:
○ Add contracts, stubs, and a stubrunner
○ Enable API back compatibility testing

So what have we achieved
149
● The Scaffold:
recommendations
● The Test:
● The Contract

Finally...

One More Thing...

Just Kidding :-)

You’re Done!

Though Really, You’re Just Beginning...
154
Because now…
● You can evolve your app faster
● With (hopefully) better code
● Consistent testing practices
● Fast failure and feedback
● Contract-based API definitions
● API and DB rollback assurance
● And…. instant pipelines from source to production!

What About Poor Simon?
155
Greeting UI Fortune
Service
Fortune DB
Simon
WHO SAID
Do What Works!?

156
Greeting UI Fortune
Service
Fortune DB
Simon
WHO SAID
Do What Works!?
Contract.make{“””
should return author name
“””)
...

157
Greeting UI Fortune
Service
Fortune DB
Simon
WHO SAID
Do What Works!?
Fortune
Service
Stub

158
Greeting UI Fortune
Service
Fortune DB
Simon
WHO SAID
Do What Works!?
V2:
ALTER TABLE fortune
CHANGE COLUMN text TO fortune,
ADD author CHAR (50);

159
Greeting UI Fortune
Service
Fortune DB
Simon
WHO SAID
Do What Works!?

160
Greeting UI Fortune
Service
Fortune DB
Simon
Pivotal says it all
the time!

Simon Spring Cloud Pipelines
161
Greeting UI Fortune
Service
Fortune DB
Oh, right!
I knew that.

Sample Code...
162
● Github repos showing the
end-state of each “act”
● Step-by-step instructions will
be published soon and made
accessible through the repo
Readmes

Customizing Spring Cloud Pipelines

Customizing Spring Cloud Pipelines
$ curl -LOk https://github.com/spring-cloud/spring-cloud-
pipelines/archive/v1.0.0.M8.zip
$ unzip v1.0.0.M8.zip
$ cd spring-cloud-pipelines-v1.0.0.M8
$ git init
$ # modify the pipelines to suit your needs
$ ./gradlew customize
$ …
$ git add .
$ git commit -a -m “Initial commit”
$ git remote add origin ${YOUR_REPOSITORY_URL}
$ git push origin master

Summary
• Continuous Deployment allows you to continuously deliver business value
• Spring Cloud Pipelines gives you OOB tooling to test your software via
• unit and integration testing
• contract testing
• rollback testing
• You can gradually migrate your applications to start using SC Pipelines
• SC Pipelines allows you to easily adjust the deployment pipeline to suit your
company’s needs
• Thanks to PaaS you can easily do A/B & zero downtime deployment

● Github Analytics: https://github.com/spring-cloud-samples/github-analytics
● Github Webhook: https://github.com/spring-cloud-samples/github-webhook
● Fortune Service: https://github.com/ciberkleid/fortune-service
● Greeting UI: https://github.com/ciberkleid/greeting-ui/
● SC-Pipelines documentation: https://cloud.spring.io/spring-cloud-pipelines/
● Pivotal Web Services trial : https://run.pivotal.io/
● PCF Dev (CF on your laptop) : https://docs.pivotal.io/pcf-dev/
Links

Learn More. Stay Connected.
▪ Read the docs
http://cloud.spring.io/spring-cloud-pipelines/
▪ Talk to us on Gitter
https://gitter.im/spring-cloud/spring-cloud-pipelines
Twitter: twitter.com/springcentral
YouTube: spring.io/video
LinkedIn: spring.io/linkedin
Google Plus: spring.io/gplus

Continuous Deployment to the cloud

More Related Content

What's hot (20)

Similar to Continuous Deployment to the cloud (20)

More from VMware Tanzu (20)

Recently uploaded (20)

Continuous Deployment to the cloud