Skip to main content
Effective capacity planning is crucial as your CI/CD workloads grow. In this guide, we explore the two primary scaling strategies—horizontal and vertical—and show how to apply them to a Jenkins environment. You’ll learn how to provision resources, distribute workloads, and ensure high availability for your Jenkins pipelines.

​
Horizontal vs. Vertical Scaling

When demand rises, you have two main options:
Scaling StrategyDescriptionKey BenefitsKey Drawbacks
Horizontal (Scale-Out)Add more servers or nodes to share the load• Fast capacity increase
• Better fault tolerance		• More infrastructure to manage
• Potential over-capacity for light workloads
Vertical (Scale-Up)Upgrade CPU, memory, or storage on an existing server• Easier single-node management
• Cost-effective for steady, small workloads		• Hardware limits
• Single point of failure

​
Horizontal Scaling (Scale-Out)

  • Quick provisioning: Spin up additional servers or containers.
  • Resilience: If one node fails, others continue processing jobs.
  • Elasticity: Automatically add or remove agents based on queue length.
The image explains horizontal scaling, showing multiple servers and listing its benefits and drawbacks. Benefits include easier scaling and improved fault tolerance, while drawbacks involve managing multiple machines and potential cost issues for smaller loads.

​
Vertical Scaling (Scale-Up)

  • Resource boost: Increase CPU cores, RAM, or disk on your master node.
  • Simplified setup: Only one instance to monitor and secure.
  • Cost savings: May be cheaper for predictable, low-traffic workloads.
The image explains vertical scaling, showing stacked servers and listing its benefits and drawbacks. Benefits include simpler management and cost-effectiveness for smaller workloads, while drawbacks include physical limits and a single point of failure.

​
Applying Scaling to Jenkins

As CI/CD adoption spreads, a single Jenkins controller can become a bottleneck. Build jobs, plugins, and user sessions all consume heap memory and CPU, leading to longer garbage-collection pauses and potential downtime.
Refer to the Jenkins System Requirements before planning upgrades or adding agents.

​
Limits of Vertical Scaling in Jenkins

  • Increasing heap size can worsen GC pause times.
  • Physical hardware upgrades hit a ceiling and require maintenance windows.
  • One controller means one failure can pause all pipelines.
Relying solely on vertical scaling creates a single point of failure—if the master node goes down, all CI/CD pipelines stop.

​
Embracing Horizontal Scaling with Build Agents

Distribute workloads by adding Jenkins agents (nodes):
  • Offload builds: Keep the controller lean by running heavy jobs on agents.
  • High availability: Agent failures don’t impact the master’s uptime.
  • Rolling upgrades: Update or replace individual agents without halting your CI/CD pipeline.
The image illustrates Jenkins scaling, showing vertical and horizontal scaling with CPU and memory resources, and highlighting issues like garbage collection, physical limits, and single points of failure. It also depicts different job statuses with weather icons.

​
Conclusion

For most growing Jenkins environments, horizontal scaling—deploying and managing multiple build agents—offers superior resilience and performance. Combine this with a scalable storage solution for artifacts and logs to complete your capacity-planning strategy.

Watch Video