Why Milestone step for concurrency in Jenkins? - Purpose & Use Cases

The Big Idea

What if your pipeline could automatically ignore old work and only focus on the latest changes?

The Scenario

Imagine you have many tasks running at the same time in Jenkins, like building and testing software. Without control, these tasks can pile up and cause confusion or waste resources.

The Problem

Manually managing which tasks run first or waiting for others to finish is slow and tricky. It's easy to make mistakes, like running too many tasks at once or letting outdated tasks continue, which wastes time and causes errors.

The Solution

The Milestone step in Jenkins helps by automatically stopping older tasks when a newer one reaches the same point. This keeps only the latest work running, saving time and avoiding conflicts without extra effort.

Before vs After

✗ Before

stage('Build') {
  // no control, all builds run
  build job: 'MyJob'
}

✓ After

stage('Build') {
  milestone(1)
  build job: 'MyJob'
}

What It Enables

It enables smooth, efficient pipelines that focus on the latest work, avoiding wasted effort and speeding up delivery.

Real Life Example

When developers push code quickly, the Milestone step ensures Jenkins only tests the newest code, stopping older tests that are no longer needed.

Key Takeaways

Manual concurrency control is slow and error-prone.

Milestone step automatically cancels outdated tasks.

This keeps pipelines efficient and focused on current work.

Practice

1. What is the main purpose of the milestone step in a Jenkins pipeline?

easy

A. To allow only the latest pipeline run to continue past a certain point

B. To trigger a new pipeline run automatically

C. To send notifications when a build fails

D. To archive build artifacts for later use

Solution

1. Step 1: Understand the role of the milestone step

   The milestone step is designed to control concurrency by ensuring only the newest pipeline run proceeds beyond the milestone point.

2. Step 2: Compare with other pipeline steps

   Other steps like triggering builds or sending notifications do not control concurrency like milestone does.

3. Final Answer:

   To allow only the latest pipeline run to continue past a certain point -> Option A

4. Quick Check:

   Milestone controls concurrency = B [OK]

Hint: Milestone lets only the newest run pass a checkpoint [OK]

Common Mistakes:

• Confusing milestone with notification steps
• Thinking milestone triggers builds
• Assuming milestone archives artifacts

2. Which of the following is the correct syntax to add a milestone step with label "BuildCheckpoint" in a Jenkins declarative pipeline?

easy

A. milestone label: 'BuildCheckpoint'

B. milestone('BuildCheckpoint')

C. milestone label = 'BuildCheckpoint'

D. milestone: 'BuildCheckpoint'

Solution

1. Step 1: Recall the correct milestone syntax

   The milestone step uses named parameters with the syntax: milestone label: 'LabelName'.

2. Step 2: Identify syntax errors in other options

   Options B, C, and D have incorrect syntax such as missing named parameter format or wrong punctuation.

3. Final Answer:

   milestone label: 'BuildCheckpoint' -> Option A

4. Quick Check:

   Correct milestone syntax uses label: 'name' [OK]

Hint: Use named parameter syntax: milestone label: 'Name' [OK]

Common Mistakes:

• Using parentheses instead of named parameters
• Using equal sign instead of colon
• Omitting the label keyword

3. Given this Jenkins pipeline snippet:

stage('Test') {
  steps {
    milestone label: 'TestStart'
    echo 'Running tests'
  }
}

// Two pipeline runs start quickly one after another

What happens to the first pipeline run when the second run reaches the same milestone?

medium

A. The first run continues normally alongside the second

B. The second run waits until the first run finishes

C. The first run is aborted and only the second run continues past the milestone

D. Both runs are aborted at the milestone

Solution

1. Step 1: Understand milestone concurrency behavior

   When multiple runs reach the same milestone label, Jenkins allows only the latest run to proceed and aborts older runs at that point.

2. Step 2: Apply this to the scenario

   The second run reaching 'TestStart' milestone causes the first run to abort at that milestone, so only the second run continues.

3. Final Answer:

   The first run is aborted and only the second run continues past the milestone -> Option C

4. Quick Check:

   Milestone aborts older runs = D [OK]

Hint: Milestone aborts older runs, keeps newest only [OK]

Common Mistakes:

• Thinking both runs continue
• Assuming second run waits
• Believing both runs abort

4. You added a milestone step in your Jenkins pipeline but it does not seem to abort older runs as expected. Which of the following is the most likely cause?

medium

A. The pipeline is missing the agent block

B. The milestone step is placed after all stages have completed

C. The milestone label is not unique and is reused in multiple places

D. The Jenkins server is offline

Solution

1. Step 1: Check milestone placement

   The milestone step is most likely placed after all stages have completed. This makes it ineffective because older runs finish before a newer run can reach the milestone to trigger aborts.

2. Step 2: Evaluate other options

   Label reuse affects log clarity but not concurrency control (ordinals matter). Missing agent prevents runs entirely. Server offline stops everything.

3. Final Answer:

   The milestone step is placed after all stages have completed -> Option B

4. Quick Check:

   Milestone placement before completion critical = A [OK]

Hint: Place milestone before critical/protected sections [OK]

Common Mistakes:

• Placing milestone after all stages
• Blaming unrelated pipeline syntax
• Assuming server offline causes milestone issues

5. You want to ensure that only the latest run of your Jenkins pipeline proceeds past the "Deploy" stage, aborting any previous runs stuck at that point. How should you use the milestone step to achieve this?

hard

A. Use milestone label: 'Deploy' only in the first stage

B. Place milestone label: 'Deploy' at the end of the Deploy stage

C. Place milestone label: 'Deploy' in every stage before Deploy

D. Place milestone label: 'Deploy' at the start of the Deploy stage

Solution

1. Step 1: Identify where to place milestone for concurrency control

   To abort older runs before deploying, place the milestone step at the start of the Deploy stage so only the newest run proceeds.

2. Step 2: Understand why other placements fail

   Placing milestone at the end or in other stages won't stop older runs before deployment. Using it only in the first stage doesn't control concurrency at Deploy.

3. Final Answer:

   Place milestone label: 'Deploy' at the start of the Deploy stage -> Option D

4. Quick Check:

   Milestone at stage start controls concurrency there [OK]

Hint: Put milestone at stage start to block old runs [OK]

Common Mistakes:

• Placing milestone too late in the stage
• Using milestone in unrelated stages
• Not placing milestone at all