Gauge

A gauge represents the value of something at a point in time. It can go up or down, which makes it suitable for measurements such as memory usage, queue depth, active connections, or temperature.

Temporal Aggregation

Temporal aggregation groups points from the same time series into time buckets.

Common temporal aggregations for gauges are:

• Avg to smooth variation within each interval
• Min and Max to understand range
• Last when you care about the latest observed value in each bucket

If a gauge reports values like 70, 75, and 80 during a 5-minute bucket:

• Avg returns 75
• Min returns 70
• Max returns 80
• Last returns the last reported value in that bucket

Spatial Aggregation

Spatial aggregation combines multiple series of the same metric.

Examples:

• sum memory usage across all pods in a namespace
• average CPU utilization across nodes in a cluster
• max queue depth across worker instances

Group by controls how the series are combined. Without grouping, all matching series are rolled into one result. With grouping, KloudMate preserves one result per selected dimension.

When Gauges Work Well

Gauges are a good fit for:

• memory usage
• active sessions or connections
• queue depth
• concurrent workers
• disk utilization