PricingPricingDocsDocsBlogBlogBook a demoBook a demoCompanyCompany (opens in a new tab)
Product
DocsAPI ReferencePricingConsole DemoBlogCompany
Start Freev3.4.7
Wave Autoscale

Min/Max Recommendation

Recommends min and max replicas for each Deployment from 7 to 14 days of its traffic, as a Performance set and a Cost set.

The challenge

At 9 p.m. the peak hits maxReplicas, and the HPA reports ScalingLimited

kubernetes.io, HorizontalPodAutoscaler v2 API and HPA walkthrough. The 9 p.m. peak is illustrative.

minReplicas and maxReplicas are fixed numbers you set when you create the HPA. minReplicas defaults to 1 Pod, and maxReplicas is a hard cap.

When the peak needs more Pods than maxReplicas, the HPA stops at the cap and reports ScalingLimited. When minReplicas is set high to be safe, those Pods run all night.

Guessed limits waste Pods at night and cap the peak; recommended limits fit the dayWithout WaveWith Waveguessed min/maxfits the day
Without Wave
A guessed min 20 and max 24 for every hour of the dayPods neededrunning Podsmax 24min 20idle Podspeak capped00:0006:0012:0018:0024:00
At 3 a.m. the service needs about 4 Pods, but min 20 keeps 20 running. At 9 p.m. it needs 30, and max 24 stops the HPA, which reports ScalingLimited.
With Wave
The recommended Min Cost 5 and Max Performance 34 fit the same dayPods neededrunning PodsMax Performance 34Min Cost 500:0006:0012:0018:0024:00
From 7 to 14 days of traffic, Wave recommends a Performance set and a Cost set. With Min Cost 5 and Max Performance 34, the quiet hours run fewer Pods and the peak fits.

Pod counts and hours are illustrative. ScalingLimited and the minReplicas default: kubernetes.io. Performance and Cost sets, 7 to 14 days of history: Min/Max Recommendations docs.

How Wave does it

Setting min and max replicas
Limits set by hand (default)Peak capped, idle Pods at night
  1. Engineer
    sets minReplicas and maxReplicas once
  2. ↓HPA spec→
    HPA
    min 20 and max 24, every hour of the day
  3. ↓9 p.m. peak→
    HPA controller
    stops at maxReplicas
  4. ↓condition→
    kubectl describe hpa
    ScalingLimited: True
Wave Min/Max RecommendationLimits from the measured day
  1. Wave Agent
    with cAdvisor, on every node
  2. ↓gRPC→
    Wave Core
    keeps 7 to 14 days of traffic
  3. ↓history→
    Wave Intelligence
    learns the 24-hour cycle
  4. ↓two sets→
    Recommendation tab
    Performance and Cost min/max, with a 24-hour heatmap
minReplicas default, maxReplicas cap, and ScalingLimited: kubernetes.io. 7 to 14 days of history, the two sets, and the heatmap: Min/Max Recommendations docs. Pod counts and hours are illustrative.
1

Learn the daily cycle

Wave reads 7 to 14 days of each Deployment's traffic and learns its 24-hour pattern (Min/Max Recommendations docs).

2

Get two sets

A Performance set keeps more headroom for spikes. A Cost set runs leaner. A 24-hour heatmap shows the peak and the quiet hours.

3

Apply and watch

Copy the values into the Deployment's Autopilot settings. A common choice is Min Cost with Max Performance. Watch scaling for about a week (Min/Max Recommendations docs).

Related in Wave Autoscale

Autopilot →Autopilot Scheduler →Read the docs →
Book a demoStart free
Product
Wave GPUAI Infrastructure Professional ServiceWaveWave AI Agent2026 Q4
Resources
DocsAPI ReferenceBlogConsole DemoPricing
Company
STCLabBook a demoContact
STCLab© 2026 Wave All rights reserved.
Terms of ServicePrivacy PolicyCookies