Skip to main content

Controller Performance Tuning

Controller performance tuning focuses on keeping Jenkins responsive, stable, and predictable as usage grows. Most performance problems come from misconfiguration, not scale.

Why Controller Performance Degrades​

Common causes:

  • Too many plugins
  • Controller running builds
  • High executor count
  • Excessive logging
  • Large job counts without folder structure

Performance issues accumulate slowly.

Controller Responsibilities (Reminder)​

Controller should handle:

  • Scheduling
  • Metadata
  • UI
  • Configuration
  • Security

Controller should not execute builds.

CPU & Memory Tuning​

Guidelines:

  • Allocate sufficient heap memory
  • Avoid over-allocating heap
  • Monitor GC pauses
  • Tune JVM options carefully

Bad JVM tuning causes UI freezes.

JVM Heap Sizing​

General guidance:

  • Small Jenkins: 4–8 GB heap
  • Medium Jenkins: 8–16 GB heap
  • Large Jenkins: 16–32 GB heap

Always observe real usage.

Disk I/O Considerations​

  • Use fast storage for JENKINS_HOME
  • Avoid network filesystems with high latency
  • Monitor disk space and IOPS

Disk bottlenecks cause slow page loads.

Plugin Impact on Performance​

  • Each plugin adds memory overhead
  • Some plugins add background jobs
  • Remove unused plugins aggressively

Plugin count directly affects startup time.

Job & Folder Organization​

Best practices:

  • Use folders to group jobs
  • Avoid flat job structures
  • Archive old jobs

Large job counts slow Jenkins.

Log Management​

  • Reduce log verbosity
  • Rotate logs
  • Avoid debug logging in production

Logs can exhaust disk and CPU.

Monitoring Performance​

Track:

  • Controller CPU usage
  • JVM heap and GC
  • Thread counts
  • Response time
  • Queue length

Tune based on data.

Common Performance Anti-Patterns​

  • Running builds on controller
  • Excessive executors
  • Massive Jenkinsfiles
  • No cleanup policies

Best Practices​

  • Controller executors = 0
  • Minimal plugins
  • Fast storage
  • Monitor continuously
  • Scale horizontally via agents

Interview Focus Areas​

  • Why controller becomes slow
  • Heap vs OS memory
  • Plugin impact on performance