Graphql-engine: Distributed Tracing

[PlamenHristov]

Have any of you guys been able to forward trace Ids from Zipkin or Jaeger or add some kind of distributed tracing on the event callbacks?

[maxpain]

+1

Instana offers a nice way to do distributed traces across all microservices in a project.

In our case, we have a lot of queries via graphql-engine and it will be cool to see SQL queries, made by hasura, in our distributed traces.

Is it possible to implement the support of this plugin in graphql-engine?
https://docs.instana.io/ecosystem/haskell/

Or maybe support of Jaeger/Zipkin/opentracing?

[leejh10003]

+1
I want to use some kind of monitoring platform like https://www.datadoghq.com!. It would be cool to profile backend for optimizing service.

[gunar]

Inspiration: https://www.apollographql.com/docs/apollo-server/federation/metrics/

[polRk]

https://www.apollographql.com/docs/graph-manager/setup-analytics/#graph-manager-reporting-endpoint

[Bessonov]

Well, I've discovered that tracing is already built-in in the cloud offering. Furthermore there is a some work in progress: https://github.com/hasura/graphql-engine/issues/5410 https://github.com/hasura/graphql-engine/pull/5624 .

[paf31]

There is some information on trace propagation here. The beta version of Hasura Cloud used custom HTTP headers but going forward, we'll be using the B3 propagation format (the PR was recently merged), so it should be possible to integrate with ZipKin relatively easily.

[Bessonov]

Hey @paf31 thank you for your response! Is there any plan make it usable with own instance? If I understand correctly, traces are stored in the database, but I'm not sure how to activate it.

[jsalva]

+1