1. Home
  2. Knowledge base
  3. Tutorials

Using APM Insights to Troubleshoot Performance Issues

Last modified: October 9, 2025

APM Insights (Application Performance Monitoring) is a powerful diagnostic tool available in the Pressable control panel. It allows you to capture detailed information about how your site executes PHP requests, helping identify slow plugins, database queries, or theme functions that may affect performance.

This feature was previously available only to Pressable staff and is now accessible to all customers. It’s designed to be safe and stable, but if your production site is under heavy load, you may wish to use it on a clone instead. Capturing traces temporarily increases PHP activity, which can have a small performance impact while monitoring is active.

Introducing APM Insights (Changelog)

Where to find APM Insights

APM Insights is located within the Pressable control panel under:

Sites → [select your site] → Performance → APM Insights

When you open the section, you’ll see an interface like this:

This view includes:

  • A dropdown menu for selecting the monitoring duration (for example, 5 minutes).
  • Buttons to Activate Monitoring, View APM Dashboard, and Copy APM Link.

Activating APM monitoring

APM captures detailed PHP trace data for a limited time. To begin monitoring:

  1. Choose a duration from the dropdown (5 to 30 minutes).
  2. Click Activate Monitoring.
  3. A notification will confirm that APM is active, including a link to view the APM Dashboard.

Once APM is running, any requests made to your site will be recorded for the selected period. For best results:

  • Keep sessions short (around 5 minutes is typically enough).
  • Reproduce the issue you want to analyze (for example, visit a slow page or complete a checkout process).

If your site experiences significant production traffic or runs resource-heavy workloads, consider testing on a clone site instead. APM adds diagnostic overhead while collecting traces, which may slightly affect response times during active monitoring.

Viewing and sharing traces

When monitoring is active (or after traces have been captured), you can access the results by clicking View APM Dashboard. This opens a new tab with the APM viewer.

The dashboard lists recorded requests, including:

  • Timestamp: When the request occurred.
  • Duration: How long it took to process.
  • Spans: The number of individual functions or operations in the trace.
  • Errors: Count of PHP or query-level errors detected.

You can click any trace to open it for detailed inspection. You can also click Copy APM Link to generate a shareable, temporary URL for developers or collaborators.

Understanding APM traces

Each trace represents a single PHP request and includes an in-depth breakdown of how WordPress executed that request.

The main elements include:

  • Summary: Displays the request path, duration, timestamp, and status (for example, SUCCESS or ERROR).
  • Latency chart: Shows the relative time spent in WordPress core (WP), database (DB), and cache (Memcache) operations. Memcache represents the built-in object caching layer in Pressable’s environment, which is distinct from page or edge caching. Object caching stores frequently used data such as database query results, transients, and other reusable data structures in memory to improve performance and reduce redundant database queries.
  • Spans tab: Lists every function, query, or action executed as part of the request.

Look for long-running spans or stages, as these are potential bottlenecks. For example:

  • Large “db” spans may indicate slow database queries.
  • Excessive “wp” time may point to inefficient PHP or plugin hooks.
  • “memcache” spans show cache lookups, which should generally be short.

Use filters to narrow down by span type or duration to identify the slowest processes.

Saving traces

APM trace data is ephemeral by design and should not be expected to persist beyond the active monitoring session. If you identify a performance issue worth investigating further, be sure to copy the relevant trace details immediately. Trace data may be unavailable once you close the dashboard or after the monitoring session concludes.

Best practices and next steps