Trace Explorer
A trace is a record of everything that happened during a single request as it travelled across your services. Each unit of work within a trace is a span. The Trace Explorer opens any trace and gives you three ways to read it, plus attribute search to find the traces you care about.
Waterfall view
Section titled “Waterfall view”The waterfall view lays out every span in the trace chronologically, with its duration, status, and parent-child relationships. It is the fastest way to see where time is spent and whether calls run sequentially or in parallel.
Flamegraph view
Section titled “Flamegraph view”The flamegraph aggregates spans by operation name. Where the waterfall shows one request in detail, the flamegraph shows where time concentrates across a call stack, making hot paths in deep stacks easy to spot.
Span list
Section titled “Span list”The span list is a flat table of every span in the trace. Use it to scan spans by operation name, inspect their attributes, and read raw span data without navigating the tree.
Span attributes
Section titled “Span attributes”Every span carries attributes that give it context. Common ones include:
service.name— the service that produced the spanhttp.method,http.url,http.status_code— for HTTP callsdb.system,db.statement— for database queriesrpc.service,rpc.method— for gRPC calls- any custom attributes you add in your own code
Attribute search
Section titled “Attribute search”Filter traces by attribute to narrow a large set to the requests that matter — for example, all traces for one http.route, one db.system, or a specific customer ID you attach as a custom attribute. Combine attribute filters with duration and status to isolate slow or failing requests.
Trace-to-log correlation
Section titled “Trace-to-log correlation”If a service emits logs through the same OpenTelemetry SDK, Osuite links each log line to its span using trace_id and span_id. From any span in the Trace Explorer you can open the logs emitted during that span without leaving the trace. See Trace correlation for how logs are matched to traces.
Related
Section titled “Related”- APM overview — golden signals, service list, and the dependency map.
- Error tracking — errors grouped across traces.
- Trace sampling — what is kept when you sample.