- Published on
A single trace shows one request. To measure how many requests complete a multi-step flow across services, you need to aggregate thousands of traces. Trace Funnels does this.
Feature
All Posts
- opentelemetry (133)
- product (42)
- observability (42)
- signoz (42)
- product-updates (38)
- tech-tutorial (30)
- tools-comparison (28)
- opentelemetry-instrumentation (21)
- distributed-tracing (17)
- javascript (16)
- log-management (15)
- jaeger (13)
- python (11)
- open-source (11)
- monitoring (10)
- tech-resources (10)
- prometheus (9)
- logging (7)
- docker (7)
- mcp (6)
- community (6)
- launch-week (6)
- kubernetes (5)
- apm (5)
- nextjs (5)
- datadog (4)
- pricing (4)
- talks (4)
- database-monitoring (4)
- ai (3)
- nodejs (3)
- events (3)
- engineering (3)
- go--golang (3)
- logs (2)
- infrastructure (2)
- newsletter (2)
feature (2)
- guides (2)
- cost-optimization (2)
- java (2)
- metrics (2)
- net (2)
- langchain (2)
- databases (2)
- openclaw (1)
- claude-code (1)
- log (1)
- mcp-servers (1)
- hiring (1)
- splunk (1)
- alternatives (1)
- development (1)
- sampling (1)
- telemetry (1)
- redis (1)
- monitoring-tools (1)
- vercel-ai-sdk (1)
- dotnet (1)
- rust (1)
- ruby (1)
- php (1)
- web-vitals (1)
- llamaindex (1)
- golang (1)
- fastapi (1)
- elixir--erlang (1)
- feature-flags (1)
- devops (1)
- software-development (1)
- continuous-delivery (1)
- traces (1)
- nginx (1)
- n8n (1)
- mysql (1)
- llm (1)
- kubecon (1)
- operations (1)
- ai--llm-observability (1)
- coding (1)
- grafana-alternatives (1)
- go (1)
- microservices (1)
- team (1)
- grafana (1)
- aws (1)
- cloud (1)
- security (1)
- node (1)
- clickhouse (1)
- json (1)
- startups (1)
- api-monitoring (1)
- sre (1)
- jokes (1)
- Published on
When you're debugging a trace and staring at a span that took 2 seconds, you need to know if that's actually slow or just normal. Span Percentile shows you the percentile inline in the trace detail view.
