Cloud agent provisioning (EC2, Azure) in Jenkins - Step-by-Step Execution

Process Flow - Cloud agent provisioning (EC2, Azure)

Start Jenkins Job

↓

Select Cloud Provider

↓

Provision EC2

↓

Configure Agent

↓

Agent Ready for Job Execution

↓

Run Job

↓

Cleanup

↓

End

This flow shows Jenkins starting a job, choosing a cloud provider (EC2 or Azure), provisioning the agent, configuring it, running the job, and cleaning up.

Execution Sample

Jenkins

pipeline {
  agent {
    label 'ec2-agent'
  }
  stages {
    stage('Build') {
      steps {
        echo 'Building on EC2 agent'
      }
    }
  }
}

This Jenkins pipeline provisions an EC2 cloud agent labeled 'ec2-agent' and runs a build stage on it.

Process Table

Step	Action	Evaluation	Result
1	Start Jenkins job	Job triggered	Job starts execution
2	Select cloud provider	Provider = AWS EC2	EC2 provisioning begins
3	Provision EC2 instance	Request EC2 instance with label 'ec2-agent'	EC2 instance launched and connected
4	Configure agent	Install Jenkins agent on EC2	Agent ready and connected to Jenkins master
5	Run build stage	Execute echo command	Output: 'Building on EC2 agent'
6	Job completes	Build stage finished	Job ends successfully
7	Cleanup	Terminate EC2 instance	EC2 instance terminated
8	Exit	No more stages	Job fully completed

💡 Job ends after build stage completes and EC2 instance is terminated.

Status Tracker

Variable	Start	After Step 3	After Step 4	After Step 7	Final
cloud_provider	null	AWS EC2	AWS EC2	AWS EC2	null
agent_status	not provisioned	provisioning	ready	terminating	terminated
job_status	not started	running	running	completed	completed

Key Moments - 2 Insights

Why does the agent_status change from 'provisioning' to 'ready' after step 4?

Because after the EC2 instance is launched (step 3), Jenkins installs and connects the agent software on it (step 4), making the agent ready to run jobs as shown in execution_table row 4.

Why is the EC2 instance terminated after the job completes?

To save costs and resources, Jenkins cleans up by terminating the EC2 instance after the job finishes, as shown in execution_table step 7.

Visual Quiz - 3 Questions

Look at the execution table, what is the agent_status after step 4?

A provisioning

B ready

C terminated

D not provisioned

💡 Hint

Check the variable_tracker table under 'agent_status' after Step 4.

At which step does the EC2 instance get terminated?

A Step 7

B Step 6

C Step 5

D Step 8

💡 Hint

Look at the execution_table row describing cleanup and termination.

If the cloud_provider was changed to Azure, which step would change in the execution table?

A Step 4

B Step 3

C Step 2

D Step 5

💡 Hint

Step 2 is where the cloud provider is selected.

Concept Snapshot

Jenkins Cloud Agent Provisioning:
- Define cloud provider in pipeline (e.g., AWS EC2 or Azure VM).
- Jenkins requests and launches cloud VM as agent.
- Agent installs Jenkins software and connects.
- Job runs on this dynamic agent.
- After job, agent VM is terminated to save cost.

Full Transcript

This visual execution shows how Jenkins provisions cloud agents on EC2 or Azure. The job starts, selects a cloud provider, provisions the VM, configures the Jenkins agent, runs the job, and then cleans up by terminating the VM. Variables like agent_status and job_status change step-by-step. Key moments include understanding when the agent becomes ready and why cleanup happens. The quiz tests knowledge of these steps and variable states.

Practice

1. What is the main purpose of cloud agent provisioning in Jenkins?

easy

A. To dynamically create and use cloud machines as Jenkins agents

B. To manually configure Jenkins agents on local machines

C. To backup Jenkins configuration files to the cloud

D. To monitor Jenkins job execution times

Solution

1. Step 1: Understand Jenkins agent provisioning

   Jenkins can create agents dynamically in the cloud to run jobs without manual setup.

2. Step 2: Identify the purpose of cloud provisioning

   Cloud agent provisioning automates creating cloud machines (like EC2 or Azure VMs) as Jenkins agents.

3. Final Answer:

   To dynamically create and use cloud machines as Jenkins agents -> Option A

4. Quick Check:

   Cloud agent provisioning = dynamic cloud agents [OK]

Hint: Cloud provisioning means automatic cloud agent creation [OK]

Common Mistakes:

• Confusing provisioning with manual setup
• Thinking it is for backup or monitoring
• Assuming it only works on local machines

2. Which of the following is the correct way to specify an EC2 AMI ID in Jenkins cloud agent configuration?

easy

A. ami-0abcdef1234567890

B. ec2-ami-123456

C. AMI:1234567890

D. image-ec2-abcdef

Solution

1. Step 1: Recognize EC2 AMI ID format

   EC2 AMI IDs start with 'ami-' followed by alphanumeric characters, e.g., 'ami-0abcdef1234567890'.

2. Step 2: Match the correct format

   ami-0abcdef1234567890 matches the standard EC2 AMI ID format exactly.

3. Final Answer:

   ami-0abcdef1234567890 -> Option A

4. Quick Check:

   EC2 AMI ID starts with 'ami-' [OK]

Hint: EC2 AMI IDs always start with 'ami-' [OK]

Common Mistakes:

• Using incorrect prefixes like 'ec2-ami-'
• Confusing AMI ID with image names
• Adding extra labels or colons

3. Given this Jenkins EC2 cloud agent snippet:

instanceCapStr: '3',
idleTerminationMinutes: 10,
ami: 'ami-0abc123def456',
remoteFS: '/home/jenkins'

What happens when 4 jobs are triggered simultaneously?

medium

A. Jenkins throws an error due to instance cap mismatch

B. 4 agents will be created ignoring the instance cap

C. All 4 jobs run on a single agent

D. Only 3 agents will be created; the 4th job waits in queue

Solution

1. Step 1: Understand instanceCapStr meaning

   The 'instanceCapStr' limits the max number of EC2 agents Jenkins can create to 3.

2. Step 2: Analyze job triggering behavior

   When 4 jobs start, Jenkins creates only 3 agents due to the cap; the 4th job waits in the queue.

3. Final Answer:

   Only 3 agents will be created; the 4th job waits in queue -> Option D

4. Quick Check:

   Instance cap limits agents = jobs queue if exceeded [OK]

Hint: Instance cap limits agents; extra jobs queue [OK]

Common Mistakes:

• Assuming Jenkins ignores instance cap
• Thinking all jobs run on one agent
• Expecting an error when cap exceeded

4. You configured an Azure cloud agent in Jenkins but agents never start. Which is the most likely cause?

medium

A. Instance cap set to zero

B. Jenkins master is offline

C. Incorrect Azure VM image name or ID

D. Idle timeout set too high

Solution

1. Step 1: Check Azure VM image configuration

   If the VM image name or ID is wrong, Jenkins cannot provision agents on Azure.

2. Step 2: Evaluate other options

   Jenkins master offline stops all jobs, but question focuses on agent provisioning. Instance cap zero means no agents, but usually agents would not start at all. Idle timeout affects termination, not start.

3. Final Answer:

   Incorrect Azure VM image name or ID -> Option C

4. Quick Check:

   Wrong VM image blocks agent start [OK]

Hint: Check VM image name first if agents don't start [OK]

Common Mistakes:

• Ignoring VM image correctness
• Blaming idle timeout for start issues
• Assuming Jenkins master offline without checking

5. You want to optimize Jenkins cloud agents on EC2 to reduce cost. Which combination of settings is best?

hard

A. Set instance cap high and idle termination to 60 minutes

B. Set instance cap low and idle termination to 5 minutes

C. Disable instance cap and idle termination

D. Set instance cap low and disable idle termination

Solution

1. Step 1: Understand instance cap effect

   A low instance cap limits max running agents, controlling cloud costs.

2. Step 2: Understand idle termination effect

   Setting idle termination to 5 minutes stops unused agents quickly, saving cost.

3. Step 3: Evaluate other options

   High cap or disabling termination increases cost. Disabling termination wastes resources.

4. Final Answer:

   Set instance cap low and idle termination to 5 minutes -> Option B

5. Quick Check:

   Low cap + quick termination = cost saving [OK]

Hint: Low cap + short idle timeout saves cloud cost [OK]

Common Mistakes:

• Setting high caps wastes money
• Disabling idle termination keeps idle agents running
• Ignoring cost impact of settings