Skip to content

eureka-avg

Input Stack:
query: Query
 ⇨ Output Stack:
TimeSeriesExpr

Compute an average based on the number of instances in the UP state according to Eureka service discovery. This operator uses the discovery.status metric to determine the active instance count as the denominator, making it ideal for metrics related to incoming traffic that is load-balanced through Eureka.

Parameters

  • query: Query for metrics to average, typically traffic-related metrics routed via Eureka

How It Works

  1. Numerator: Uses the total from the provided query (typically aggregated metrics)
  2. Denominator: Counts instances with state=UP for the same service
  3. Scoping: Common infrastructure tags automatically restrict the denominator scope
  4. Result: Average per active (traffic-receiving) instance

When to Use

This operator is specifically designed for scenarios where:

  • Metrics represent incoming traffic routed through Eureka service discovery
  • You want average per active instance, not per total deployed instance
  • The goal is to measure per-node load for traffic-bearing services

Examples

Computing average requests per second per active instance:

name,sps,:eq,
nf.app,nccp,:eq,
:and,
:eureka-avg

Comparison with Other Averages

Operator Denominator Use Case
:avg Number of reporting time series General averaging
:node-avg Number of unique instances Per-instance average (all instances)
:eureka-avg Number of UP instances in Eureka Per-active-instance average (Eureka-routed traffic)
:dist-avg Recorded sample count Per-measurement average

Infrastructure Tag Scoping

The operator automatically uses common tags to ensure the denominator matches the numerator scope:

  • nf.app - Application name
  • nf.cluster - Cluster identifier
  • nf.region - AWS region
  • Other standard Netflix infrastructure tags
  • :avg - Standard average across time series
  • :node-avg - Average using total instance count (includes non-traffic instances)
  • :dist-avg - Average using measurement sample counts
  • :sum - Total aggregation (numerator for manual average calculation)