Profiles catalog

View every service in your environment — including services that do not emit spans or metrics — from the Profiles tab in the APM Service Catalog. Profiling data alone is enough to surface a service, giving you full coverage of runtime behavior alongside the rest of your APM workflows.

Why use Profiles catalog

Continuous profiling gives engineers code-level visibility into CPU and memory usage with minimal overhead — learn more about how Coralogix empowers its users with profiling data in our continuous profiling documentation. Profiles catalog makes this data even more powerful by embedding it directly in the Service Catalog, unifying profiling with the rest of your APM workflows.

• Single source of truth for service performance: Profiling data appears directly in the Service Catalog, alongside spans, metrics, logs, and Service Level Objectives (SLOs). Teams no longer need to pivot between tools or dashboards to understand CPU and memory usage across services.

• Visibility into all services, not just instrumented ones: The catalog includes profiled services even if they lack traces or span metrics, ensuring full coverage of runtime behavior across your environment.

• Faster triage and root-cause analysis: From Service Catalog, engineers can see which services are the most resource-intensive and drill down into function-level profiles—all without leaving the APM workflow.

• Trend discovery in context: Track performance regressions or optimization opportunities over time, by service, version, or environment.

• Built-in profile comparison for faster insights: Compare profiles across time ranges, versions, or environments directly from the catalog. Visual diffs in flame graphs highlight regressions, improvements, and code paths unique to each profile, making it easy to validate fixes and catch performance issues early.

Key capabilities

Filter and reset the catalog

The catalog toolbar exposes three filters that scope which services appear: a time-range picker, an environment selector, and a service-name search. The selections persist in the page URL so refreshing or sharing the link preserves the view.

Select Reset to clear all three filters in one action:

• The time range returns to the last 15 minutes.
• The environment selector reselects every environment available for the active time range.
• The service-name search clears.

Unified service visibility

• Profiles catalog displays all services in your environment.
• This includes:
  • Services with profiling data only (no traces or metrics).
  • Services with full APM instrumentation (spans, metrics, and profiling).

Switch between CPU and memory views

The catalog header shows two chips — CPU and Memory — that switch the catalog between profile types. Each chip carries a badge with the number of services that have data for that profile type under the current time-range and environment filters, so you can see coverage at a glance without switching. Switching between CPU and Memory is instant — the catalog loads both views upfront on the initial page load.

The catalog columns swap to match the selected chip:

• CPU view — Service, Profile status, Last profiled, Avg. CPU cores, Total CPU time, Samples per second, Avg. wall, Wall samples per second. The Avg. wall sparkline and Wall samples per second count show the wall-clock profile alongside the CPU columns.
• Memory view — Service, Profile status, Last profiled, Avg. memory allocations, Allocated bytes, Allocations count, Avg. heap bytes, Heap bytes, Heap objects. The Avg. heap bytes sparkline, Heap bytes, and Heap objects columns show heap data alongside the allocation columns. Byte values render in byte units (KB/MB/GB); object counts render in quantity units (K/M/B).

Read each row

• Each row shows a profile status that depends on the selected timeframe:

  • Active: Profiling data was sent within the chosen timeframe.
  • Inactive: No profiling data was sent during the chosen timeframe.

• The Last profiled column shows when data was last received for the service.

• The remaining columns are the metric columns for the selected profile type: direct performance signals you get without span or metric correlation.

Drill down into a service profile

Selecting a service in the Profiles tab opens a dedicated drilldown for that service.

• Two cards at the top show the available profile types, each with a variant dropdown:
  • CPU card — mCore (CPU profile) and Wall (wall-clock profile).
  • Memory card — Allocated bytes, Allocated objects, Heap bytes, and Heap objects.
• Each dropdown option is enabled only when that profile type has data; the card is disabled only when no option in its family has data.
• Selecting an option switches the Profile usage card below to that profile type. The selection persists in the URL across explorer and comparison modes.
• The Profile usage card adapts to the selected type:
  • CPU and Wall: Metrics chart titled CPU Cores, with a Cores / Time unit toggle. Flame graph Color by option: Total CPU time.
  • Memory: Metrics chart titled Memory Allocations, with a Rate / Total unit toggle. Flame graph Color by option: Total allocations.
• Below the chart, the Profiles section offers split, flame, and table view modes, plus a Color by dropdown to recolor the flame graph.
• Select Analyze above the flame graph to get an AI-powered breakdown of what the profile shows — see Explain flame graph for what the analysis covers.

If the selected profile type has no data, the view falls back to the first available type in the same family:

• CPU family: priority CPU → Wall. Entering the drilldown from the catalog with CPU selected: if CPU has no data and Wall does, the view switches to Wall; if neither has data, it stays on CPU.
• Memory family: priority Allocated bytes → Allocated objects → Heap bytes → Heap objects.

Profile comparison and change analysis

Profiles catalog enables side-by-side comparison of CPU and memory profiles for the same service across different timeframes, versions, or environments. You compare a baseline profile (A) with a newer profile (B) to:

• Validate performance improvements.
• Detect CPU or memory regressions early.
• Identify functions that appear only in one profile.

Visual diffs in flame graphs make changes immediately visible, helping teams understand what changed and where without manual analysis.

Guided tour

When you first visit the Profiles tab, an onboarding modal introduces the key profiling capabilities — including service drilldowns, profile comparison, profile trend visualization, and view type switching. Select Start tour to follow the walkthrough, or dismiss it to explore on your own.

Usage example

1. Select APM, then Service Catalog.
2. Switch to the Profiles tab.
3. Select the CPU or Memory chip in the catalog header to choose the profile type.
4. Select a service to open its profiling drilldown, where you can explore the chosen profile type, function-level breakdowns, and flame graphs.

The Profiles tab lists every profiled service with its profile status, last-profiled timestamp, and the metric columns for the selected view.

Additional resources

Next steps

Learn how to provide human-readable profiling insights for native languages by uploading debug symbols.