Description
Hi,
when running the adapter in my cluster I often get OOM-Kills because of a runaway resource consumption.
It would be nice if prometheus-adapter would honor configured resource limits, so it is more stable to operate.
The simplest way would be to use https://github.com/uber-go/automaxprocs for CPU limits and https://github.com/KimMachineGun/automemlimit for memory resource limits.
Depending on how much configurability is required (both default to NOOP if no (cgroup) limits are configured) the integration complexity ranges form two _-imports to adding three flags (enabling each feature + configuring percentage of memory quota used for go process)
I'd be happy to add the required functionality, but I'd first like to see what meets the project requirements.