Seeking Guidance on Implementing OpenTelemetry(for tracing), ClickHouse, and Kubernetes for Trace Visualization in a Microservice Architecture

I’m currently embarking on a journey to enhance the observability of my Node.js microservice application running on Kubernetes. After thorough research, I’ve got some key decisions to make and I would greatly appreciate advice and insights on.

  • I want to use OpenTelemetry(auto-instrumentation) for getting traces of my node.js microservices.
  • I want to use clickhouse for time-series data collection.
  • I have my application running on kubernetes cluster along with istio(for traffic routing) setup.
  • I have around 100 nodes and 35-40 Microservices.
  • We are using backend technologies like kafka, Cockroachdb, pinot, redis, Solr, elasticsearch.
  • And I do not want to use any proprietary tools like Groundcloud,New Relic,Datadog.

I need your guidance that, which open-source tool I should use for trace visualization. I have some in my mind(grafana tempo, Zipkin, jaeger).

So i need your suggestions, what components and tools should i use or need to use, which technology stack would be good fit as per my requirement.

I would recommend combination: Grafana, Tempo, Beyla.

Hi jangaraj, thanks for your response.

I would greatly appreciate it if you can tell me the flow to setup or any guidance and how it will work, it would be great.