Jaeger
The CNCF graduated distributed tracing system, originally from Uber. The default open-source choice for traces.
| License | Apache-2.0 |
| Repository | https://github.com/jaegertracing/jaeger |
| GitHub stars (approx) | 23000 |
| Language | Go |
| Self-host deployment | docker, kubernetes (jaeger-operator) |
| Commercial hosted option | no / self-host only |
| Freshness status | Fresh |
| Freshness checked | 2026-05-17 |
| Categories | traces |
| Replaces SaaS | datadog, new-relic, honeycomb |
Jaeger is the boring, dependable answer to “we need distributed tracing.” CNCF graduated, OpenTelemetry-native ingest, multiple storage backends (Cassandra, Elasticsearch, OpenSearch, ClickHouse via v2).
Pick it if: you want a battle-tested traces-only backend that plays nicely with OpenTelemetry collectors and you already operate Cassandra/ES/OpenSearch.
Skip it if: you also need metrics and logs in the same pane (see SigNoz or Tempo+Grafana stack). Jaeger UI is functional, not delightful — most teams view traces through Grafana with Jaeger as the datasource.
Deployment: Helm chart and operator for k8s; docker-compose for evaluation. Jaeger v2 is OTel-collector-based and is the current direction.