👋 Hello Team, I’ve just started experimenting with

OpenTelemetry

for one of my new projects, so I might be asking something a bit naive. Could you help me understand the differences between using the standard opentelemetry-collector and the signoz-otel-collector? Also, wouldn’t it be more vendor-agnostic to use an exporter to send telemetry data to the SigNoz backend instead of using the SigNoz Collector? Thanks!

There's a standard OTel wire schema for traces, logs, metrics which is

opentelemetry-collector

. That's the one that would run nearest to your apps. It even has a generic Clickhouse exporter for putting that data into a DB so it's queryable by tools like Grafana. But for building special UI features like Signoz, a custom database schema is needed and that's what

signoz-otel-collector

does even though it still uses Clickhouse: https://github.com/SigNoz/signoz-otel-collector/commit/a17d0fc65a2e454fce411262a099d32f444891f0 If you use the k8s operator with sidecars, your stack ends up being:

app -> opentelemetry-collector -> signoz-otel-collector -> clickhouse

Thanks @Hien Le for the explanation

a custom database schema is needed and that's what

signoz-otel-collector

does even though it still uses Clickhouse

Just curious - schemas and everything signoz specific is part of Signoz Platform, and the Platform can still do those things once the data is received from the standard exporter(otel-collector)...no?