13 Apr 2021 12:37 AM - last edited on 10 May 2022 04:37 AM by MaciejNeumann
It seems that a new (Docker) container within a cluster needs a restart to enable oneAgent to inject.
(it is dynamically linked (-; so no worries there)
An explanation is that the Go process is started before the OneAgent.
Does anybody have a remedy for this?
In my situation the OneAgent is installed on the OS,
This is just a node exporter for Prometheus. Do you really need to instrument it?
Short answer is no, on most instances this is turned off (-; but I am puzzled about the why,, I was assuming that processes within new containers will be automatically monitored, when the Agent is installed. But in this scenario it looks like the exporter is started before the agent initialization?
Thanks for your reply,