Open anovv opened 7 months ago
The engine should report various metrics, e.g. tx rx per operator, cpu/mem usage, disk/network io; label per-operator, per-job, per-cluster, expose via PromClient?
The engine should report various metrics, e.g. tx rx per operator, cpu/mem usage, disk/network io; label per-operator, per-job, per-cluster, expose via PromClient?