Process Exporter Metrics

This page documents Prometheus metrics collected from the Process Exporter, which reports resource usage statistics for grouped processes running on Yellowbrick nodes.

Purpose

These metrics track system-level resource consumption by named process groups. They are used to:

• Monitor CPU and memory utilization across core services
• Track open file descriptors, thread counts, and process sprawl
• Analyze I/O throughput (read/write bytes)
• Identify resource leaks, runaway processes, or unexpected restarts

The metrics are grouped using label groupname, allowing for fine-grained service-level observability.

Metrics

Name	Type	Freq	Labels	Description
namedprocess_namegroup_cpu_seconds_total	counter	10s	groupname	CPU seconds used by process group
namedprocess_namegroup_memory_bytes	gauge	10s	groupname	Memory used by process group
namedprocess_namegroup_num_procs	gauge	10s	groupname	Number of processes in group
namedprocess_namegroup_num_threads	gauge	10s	groupname	Threads used by process group
namedprocess_namegroup_open_filedesc	gauge	10s	groupname	Open file descriptors
namedprocess_namegroup_read_bytes_total	counter	10s	groupname	Bytes read by process group
namedprocess_namegroup_write_bytes_total	counter	10s	groupname	Bytes written by process group