tobs
tobs copied to clipboard
timescale
Spike: Enable tracing on all components in the stack
What is missing?
Since promscale is able to ingest tracing data, we should enable tracing on all components shipped with the stack. Data should be sent to opentelemetry-collector which will then forward it to promscale.
Why do we need it?
To increase adoption of the stack and allow better debugging options.
Anything else we need to know?:
Unless we think this is a must for more visibility into the performance / health of the Kubernetes cluster and the different components we deploy, I would postpone to a future epic.
Considering the effort involved. I second Ramon, If possible I would recommend for a quick PoC to evaluate the value proposition to the users.
This issue is not about kubernetes, but about gathering tracing data from components shipped with tobs like promscale, prometheus (Issue about kubernetes tracing data is in https://github.com/timescale/tobs/issues/395.). Basically it allows our APM dashboards to have data from the start. Additionally it is important for us to have visibility into our stack running the promscale data lake on cloud. As such I am against removing this issue from epics.