Use metrics and other resources to improve and optimize Puppet Enterprise performance – Puppet Support

Learn to use metrics and other resources to improve and optimize Puppet Enterprise (PE) performance, when you first install PE, when you're thinking of making changes to your infrastructure, or after you notice performance issues.

Use the puppet infrastructure tune command to get optimized settings
Use metrics and Java tools to find and fix Puppet Enterprise performance issues
Learn about scaling

Version and installation information

PE version: All supported
Note: We can't provide support for third-party software.

Solution

Use the `puppet infrastructure tune` command to get optimized settings

When you first start using Puppet Enterprise, or when you're thinking of making changes to your infrastructure, use the puppet infrastructure tune command to help you get suggested configurations to optimize available system resources and performance in Puppet Enterprise.

Use metrics and Java tools to find and fix Puppet Enterprise performance issues

Metrics tools help you get information to improve performance issues.

Metrics basics

Diagnose issues with Puppet services using the puppetlabs-puppet_operational_dashboards module, which configures Telegraf, InfluxDB, and Grafana to collect, store, and display metrics collected from Puppet services.

You can troubleshoot and fix PE performance issues using the puppetlabs-puppet_metrics_collector module (built into PE).

If you're using Splunk and PE, you can use the Splunk-HEC integration to visualize data from Puppet reports, the puppet_metrics_collector, and pe_status_check modules. Learn more about the puppetlabs-splunk_hec module for PE and the Puppet Report Viewer.

Get Puppet DB metrics from the PuppetDB dashboard

PuppetDB hosts a performance dashboard on port 8080 with performance information and metrics, including memory use, queue depth, command processing metrics, duplication rate, and query stats.

Use Puppet Profile Parser for Puppet Server issues

Graph and investigate Puppet Server performance issues with the Puppet Profile Parser

Get metrics from APIs

You can also use APIs to gather metrics and performance information.

Get JRuby, HTTP route, and catalog compilation profiler metrics from the metrics endpoint.

Measure how long it takes for Puppet Server to make requests to and receive responses from other Puppet using HTTP client metrics for Puppet Server

Get advanced Puppet Server and health metrics

Use Java tools to get information about performance issues

Puppet Server and PE services run on the JVM (Java virtual machine). When PE is unresponsive or has performance issues, you can use built-in tools in Java to get troubleshooting information.

Learn about scaling

Once you've been up and running for a while, you might need to scale to improve performance or for other organizational reasons. Get information about scaling from our documentation.