3 ms·

I love the concept ( and have implemented similar systems to the one listed ), but I am not seeing anything other then a unification of the lexicon used to talk

by dkhenry 11y ago

I love the concept ( and have implemented similar systems to the one listed ), but I am not seeing anything other then a unification of the lexicon used to talk about tracing user space programs. You still need to solve the biggest problem yourself ( i.e. Passing around context so you can have a unified trace ). So while the API might be nice to target without providing the other bits I can't see why I would adopt this over something completely custom to my environment.

joeblubaugh 11y ago
I believe the idea is that you can now build better analysis tools on top of this unified semantics - this better decouples tracing systems from the trace data analysis tools.
dkuebric 11y ago
This is definitely a concern--in fact, it's one that we need to solve in order to have distributed tracing work at all. A standard for in-process context propagation is not planned for OT1.0rc spec, but I'm a big proponent of including that down the road. It sounds like you may have some experience and/or interesting use-cases, if so, happy to chat sometime, or feel free to drop by gitter/github and join in the discussion.