add process_cpu_seconds_total to metrics

[sed-i]

It would be handy if avalanche included process_cpu_seconds_total in its metrics.

Example from prometheus/alertmanager/grafana metrics:

# HELP process_cpu_seconds_total Total user and system CPU time spent in seconds.
# TYPE process_cpu_seconds_total counter
process_cpu_seconds_total 1.61