Canary deployment pattern in Jenkins - Step-by-Step Execution

Process Flow - Canary deployment pattern

Start Deployment

↓

Deploy Canary Version

↓

Monitor Canary

↓

Promote

↓

Deploy Full

↓

End Deployment

This flow shows deploying a small subset (canary), monitoring it, then either promoting to full or rolling back based on results.

Execution Sample

Jenkins

pipeline {
  agent any
  parameters {
    booleanParam(name: 'canaryOk', defaultValue: false, description: 'Indicates if canary deployment is healthy')
  }
  stages {
    stage('Deploy Canary') {
      steps { echo 'Deploying canary version' }
    }
    stage('Monitor Canary') {
      steps { echo 'Monitoring canary health' }
    }
    stage('Promote or Rollback') {
      steps {
        script {
          if (params.canaryOk) {
            echo 'Promoting to full deployment'
          } else {
            echo 'Rolling back canary deployment'
          }
        }
      }
    }
  }
}

A Jenkins pipeline that deploys a canary, monitors it, then promotes or rolls back based on a parameter.

Process Table

Step	Action	Condition	Result	Output
1	Deploy Canary stage runs	N/A	Canary deployed	Deploying canary version
2	Monitor Canary stage runs	N/A	Canary monitored	Monitoring canary health
3	Check canaryOk parameter	canaryOk == true	Promote full deployment	Promoting to full deployment
4	End pipeline	N/A	Success	Pipeline completed
5	Alternative check	canaryOk == false	Rollback canary	Rolling back canary deployment
6	End pipeline	N/A	Failure handled	Pipeline completed with rollback

💡 Pipeline ends after promotion or rollback based on canaryOk parameter.

Status Tracker

Variable	Start	After Step 3	Final
params.canaryOk	undefined	true or false (set externally)	true or false (used to decide)

Key Moments - 3 Insights

Why do we deploy only a small part of the system first?

Deploying a small canary limits risk. If issues occur (see step 5 in execution_table), we rollback quickly without affecting all users.

What happens if the canary monitoring fails?

If monitoring shows problems (canaryOk == false), the pipeline rolls back the canary (step 5), preventing faulty code from reaching full deployment.

How does the pipeline decide to promote or rollback?

It uses the parameter 'canaryOk' checked in step 3. True promotes full deployment, false triggers rollback.

Visual Quiz - 3 Questions

Look at the execution_table, what output appears when the canary is healthy?

A Promoting to full deployment

B Deploying canary version

C Rolling back canary deployment

D Monitoring canary health

💡 Hint

Check row 3 in execution_table where canaryOk == true.

At which step does the pipeline decide to rollback the canary?

A Step 3

B Step 5

C Step 2

D Step 1

💡 Hint

Look for rollback action in execution_table rows.

If the canaryOk parameter is false, what is the final pipeline output?

A Pipeline completed

B Promoting to full deployment

C Pipeline completed with rollback

D Deploying canary version

💡 Hint

See the last row in execution_table for canaryOk == false.

Concept Snapshot

Canary deployment pattern in Jenkins:
- Deploy a small subset (canary) first
- Monitor canary health
- If healthy, promote to full deployment
- If issues, rollback canary
- Use parameters to control flow
- Minimizes risk by limiting exposure

Full Transcript

The canary deployment pattern in Jenkins starts by deploying a small part of the new version called the canary. Then the system monitors this canary for any problems. If the canary is healthy, the pipeline promotes the deployment to all users. If problems are detected, the pipeline rolls back the canary to avoid impacting everyone. This decision is controlled by a parameter called canaryOk. This approach reduces risk by testing changes on a small scale before full rollout.