Hi Team,
I want to dump thread info or different memory usage by apollo router to New Relic?
Is it supported.
My use-case is to debug a pod/container in terms of threads or memory when It has been killed on prod env.
Is there any other APM that I can use for apollo router to get this info?
I tried prometheus, but kubernetes doesn’t support thread info dumping into prometheus?
Also Is there any tool that Can I use for a live pod for memory and thread dump info just like heapdump and thread dump for java?
Hey @Ajay_Berwal, your question about getting resource information out of the runtime is a good question but there is nothing about the Router app or config itself that has insights to the memory/CPU/threads to provide o11y data.
This would have to be a feature of your runtime platform which is agnostic to Router.
If you want to go to the pod level, you can use a tool like the GNU Debugger (gbd) to build a debuggable Router binary
CARGO_PROFILE_RELEASE_DEBUG=true cargo build --release
and start the router then run the following on the pod:
sudo gdb -batch -ex "thread apply all bt" -p <router pid>
You can also access the memory and CPU usage as meters from K8s and export those to Prometheus or Otel if your runtime supports that