nb_cfg metrics will be very useful on visualizing OVN health status on the verge of scaling, we encountered compute agents begin to get dead after increasing our nodes beyond 160... southbound DB started to get slow caused this, a monitoring system on these metrics will enable us to understand and monitor these conditions...
I recommend to add these metrics to the exporter:
ovn_nb_global{segment=nb_cfg} 313223
ovn_nb_global{segment=sb_cfg} 313223
ovn_nb_global{segment=hv_cfg} 313223
ovn_chassis_private{chassis="compute2001"} 313223
nb_cfg metrics will be very useful on visualizing OVN health status on the verge of scaling, we encountered compute agents begin to get dead after increasing our nodes beyond 160... southbound DB started to get slow caused this, a monitoring system on these metrics will enable us to understand and monitor these conditions...
I recommend to add these metrics to the exporter: