Helipod MCPConnect
Helipod
PRICINGBLOG
See All Features→
TemplatesDeploy ready-made apps in one clickDocumentationAPIs, MCP and complete Helipod guidesMCP & AI AgentRun Helipod from Claude Code, Codex and CursorCommunityJoin the Telegram communityChangelogLatest updates and new features
AffiliateEarn a commission on every referralPartnershipWork together with Helipod
Language
Enter Cockpit
Horizontal Pod Autoscaler

Autoscaling Kubernetes Can Breathe.

HPA monitors CPU usage and adjusts replica count in real-time. You focus on code, Helipod handles infrastructure.

Starting from Rp 1,100/day for 1 replica. Auto-scales based on traffic.

Start NowView Pricing
Horizontal Pod AutoscalerDemo
api-helipodmain ← helipodapp/api-helipod · #1
Running
api-helipod.helipod.app
2 vCPU 2GB
1
api-helipodmain ← helipodapp/api-helipod · #2
Running
2 vCPU 2GB
1
api-helipodmain ← helipodapp/api-helipod · #3
Running
2 vCPU 2GB
1
api-helipodmain ← helipodapp/api-helipod · #4
Running
2 vCPU 2GB
1
CPU Usage45%
Target: 70%1 replica
Stable
Min: 1Max: 10

Why it matters

What you get

Target CPU %

Set CPU usage threshold (default 70%). When average CPU exceeds the target, HPA automatically adds replicas.

Scale Up

Traffic rising? HPA adds replicas as soon as the target is exceeded, without you stepping in. When it's quiet, replicas are removed again.

Scale Down

Traffic dropping? HPA gradually reduces replicas. Your balance stays safe, no resources wasted.

Min/Max Replicas

You set the limits. Minimum 1 pod always running, maximum based on budget. HPA will never exceed the cap.

Steps

How it works

  1. 01

    Enable HPA

    Open the Settings > Scale tab in your service. Enable HPA and set min/max replicas.

  2. 02

    Set Threshold

    Set target CPU usage (default 70%). You can adjust based on your application needs.

  3. 03

    Let HPA Work

    HPA monitors CPU usage every few seconds and adjusts replica count automatically.

Blog

Read also

Guides and tutorials related to this feature.

Horizontal Pod Autoscaler (HPA) - Auto-scale Pods Based on Traffic

Read

Automatic Autoscaling with HPA: Scale CPU and RAM When Traffic Rises

Read

FAQ

Frequently asked questions

Does HPA require at least 2 replicas?
Yes, HPA requires at least 2 replicas to work. This is because HPA performs horizontal scaling, adding or reducing the number of pods.
How long does scaling take?
Kubernetes HPA evaluates metrics periodically (every 15 seconds by default). New replicas are ready once the container starts and passes its health check, so the timing depends on your app.
Are there additional costs for HPA?
No additional costs. You only pay for resources used by active replicas. When scaling down, costs automatically decrease.

More features

Works even better together

Custom DomainConnect your own domain to each service. Automatic HTTPS without manual SSL setup.Internal VPCService-to-service traffic over internal DNS, without opening ports to the internet.Live LogsView logs from all pods and replicas in real-time directly from your browser.Dashboard CustomizationArrange metric blocks with drag-and-drop and pick the time range you need.Terminal AccessOpen a shell into a running container from your browser. No SSH key or VPN.Heli Crew AIManage Helipod from Claude, ChatGPT, Codex or the dashboard chat. It reads logs and runs deploys or restarts after you approve.
View All Features

Ready to try?

Sign in with Google or GitHub, connect a repository, and pay from a Rupiah balance. No credit card needed.
Start NowView All Features
WhatsApp