Workload Manager Metrics

This page documents Prometheus metrics related to the Workload Manager (WLM), which governs resource allocation and query concurrency across pools in Yellowbrick.

Purpose

The Workload Manager ensures fair and efficient use of system resources across multiple query workloads. This metric allows operators to:

• Track how many queries are currently running or waiting in each resource pool
• Identify pool-level contention or saturation
• Detect imbalances between configured concurrency limits and actual usage
• Optimize pool configurations based on query patterns

This metric is especially useful for SREs and platform admins monitoring SLA compliance or debugging queue buildup.

Metrics

Name	Type	Freq	Labels	Description
yb_lime_resourcepool	gauge	10s	cluster, cluster_name, pool, state	Query counts by running/waiting state