Continuous Profiling
Heap memory profiles and wall-clock profiling
Heap memory profiles — Investigate live retained memory in your Java services with the new Heap bytes and Heap objects profile types, available alongside Allocated bytes and Allocated objects in the service drilldown and the Profiles catalog. The Profiles catalog memory view also adds Avg. heap bytes, Heap bytes, and Heap objects columns. Learn more
Wall-clock profiling — Find latency that does not show up in CPU profiles. The CPU card on the service drilldown now offers a Wall variant alongside mCore, and the Profiles catalog CPU view exposes new Avg. wall and Wall samples per second columns. Learn more
Right-click context menu on flame graph rectangles
Right-click any frame in the flame graph on the service-drilldown Profiles tab to open a context menu with four per-frame actions. The menu is available in both Explorer and Compare modes. Learn more
Show only user code toggle for flame graph and stack-trace table
Filter the flame graph and the stack-trace table on the service-drilldown Profiles tab to your own application code with the new Show only user code toggle, next to the Color by dropdown. The toggle is on by default, so library, runtime, native, and kernel frames are hidden from view — letting you focus on actionable code paths the moment you open a profile. The filter applies in Explorer and Compare modes, and to both CPU and memory profiles. Learn more
Memory profiling for Java services
Profile memory allocations in Java services using the Coralogix Continuous Profiler SDK and Async Profiler. The Profiles catalog and service drilldown now support a Memory profile type alongside CPU. Learn more
Continuous Profiling on AWS ECS-EC2
Enable Continuous Profiling for workloads running on AWS ECS clusters backed by EC2 container instances directly from the AWS ECS-EC2 OpenTelemetry onboarding flow. Learn more
Olly mini compare analysis and frame classification
Analyze compare flame graph with Olly mini — Use AI-powered analysis in Compare mode to understand performance differences between two profiles without manually inspecting every function. Learn more
Color by frame classification — Color flame graph nodes by classification (User Code, Std Lib, External Lib, GC, Native, Kernel) to see at a glance where CPU time is spent across different code layers. Learn more
Guided onboarding tour — A guided tour now introduces key profiling capabilities when you first visit the Profiles tab. Learn more
Introducing Continuous Profiling
Introducing continuous profiling — Built on eBPF and powered by OpenTelemetry, Continuous Profiling captures high-frequency stack traces, CPU cycles, memory allocations, and more—without interfering with your application. Learn more