Kubernetes Pod Autoscaling: HPA, VPA, and KEDA Explained

One of Kubernetes' most powerful features is its ability to automatically scale workloads based on observed metrics. Choosing the right scaling mechanism depends on whether your bottleneck is parallelism (HPA), resource allocation (VPA), or external event queue depth (KEDA).

Horizontal Pod Autoscaler (HPA)

HPA scales the number of pod replicas based on CPU or custom metrics. When CPU utilization exceeds the target, new pods are added; when it drops, pods are removed.

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: mirahlabs-api-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: mirahlabs-api
  minReplicas: 2
  maxReplicas: 20
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 65

Vertical Pod Autoscaler (VPA)

VPA adjusts the CPU and memory requests of existing pods based on historical usage. Useful for right-sizing pods that were initially over-provisioned. Note: VPA and HPA should not target the same metric simultaneously.

KEDA: Event-Driven Autoscaling

KEDA (Kubernetes Event-Driven Autoscaling) scales deployments based on external event sources: queue depth in SQS/RabbitMQ, Kafka consumer lag, Prometheus metrics, or HTTP request rate. It can scale to zero when idle, making it ideal for batch processing workloads.

triggers:
  - type: rabbitmq
    metadata:
      queueName: image-processing
      queueLength: "20"  # scale up when queue > 20 messages

Production Terraform & Docker Infrastructure Config

To implement this in production, here is a complete Terraform configuration template for deploying highly available target group services with auto-scaling alerts, alongside a multi-stage optimized Docker file:

# Terraform Provider AWS declaration
provider "aws" {
  region = "us-east-1"
}

# Auto Scaling Group configuration
resource "aws_autoscaling_group" "app_asg" {
  name_prefix         = "mirahlabs-app-asg-"
  desired_capacity    = 2
  max_size            = 10
  min_size            = 2
  vpc_zone_identifier = ["subnet-12345", "subnet-67890"]

  launch_template {
    id      = aws_launch_template.app_lt.id
    version = "$Latest"
  }

  target_group_arns = [aws_lb_target_group.app_tg.arn]

  tag {
    key                 = "Environment"
    value               = "Production"
    propagate_at_launch = true
  }
}

# Dynamic Scaling Policy based on Target CPU Utilization
resource "aws_autoscaling_policy" "cpu_scaling" {
  name                   = "target-cpu-scaling"
  autoscaling_group_name = aws_autoscaling_group.app_asg.name
  policy_type            = "TargetTrackingScaling"

  target_tracking_configuration {
    predefined_metric_specification {
      predefined_metric_type = "ASGAverageCPUUtilization"
    }
    target_value = 65.0
  }
}

And here is the corresponding multi-stage production Dockerfile to build lightweight, secure images:

# Stage 1: Build dependencies
FROM python:3.11-alpine AS builder
WORKDIR /app
RUN apk add --no-cache gcc musl-dev libffi-dev g++ postgresql-dev
COPY requirements.txt .
RUN pip install --user --no-cache-dir -r requirements.txt

# Stage 2: Final lightweight image
FROM python:3.11-alpine
WORKDIR /app
RUN apk add --no-cache libpq
COPY --from=builder /root/.local /root/.local
COPY . .
ENV PATH=/root/.local/bin:$PATH
EXPOSE 5001
USER 1001
CMD ["gunicorn", "--bind", "0.0.0.0:5001", "run:app"]

Production Trade-offs & Implementation Decisions

Deploying this solution in production environments requires a careful analysis of the trade-offs involved. For instance, focusing purely on consistency (such as ACID compliance) can limit network throughput and horizontal scalability. On the other hand, adopting an eventual consistency model can lead to dirty reads and requires complex conflict resolution strategies in the application layer.

At MirahLabs, our engineering teams balance these architectural constraints by separating critical transaction paths from analytics workloads. We apply message-driven architectures with idempotent consumer systems to guarantee that network failures or retries do not result in double processing or state contamination.

Real-World Benchmarks & Resource Planning

Below is a typical performance comparison profile compiled by our engineering team in staging environments under simulated loads (10k concurrent virtual users):

When capacity planning, we recommend scaling out horizontally using containerized workloads rather than vertically upgrading underlying instance models. This maximizes uptime and provides cost efficiency through dynamic scaling policies.

Security Considerations & Vulnerability Mitigations

No production blueprint is complete without addressing security. Ensure that all data paths utilize encryption in transit (TLS 1.3) and at rest (using AES-256). Furthermore, implement strict Role-Based Access Control (RBAC) to limit operations. For APIs, always enforce rate limits (e.g. using token bucket algorithms in Redis) and run continuous static application security testing (SAST) in your CI pipeline.

How MirahLabs Applies This in Practice

Our experience building high-volume solutions like MirahCare.ai and Ayurveda.ai has taught us that early optimization is often a trap, but ignoring structural security and data design early leads to fatal development blocks. We design all client products from day one to support modular extensions, robust query indexing, and standard schema definitions, ensuring rapid iteration without technical debt growth.

Production Terraform & Docker Infrastructure Config

To implement this in production, here is a complete Terraform configuration template for deploying highly available target group services with auto-scaling alerts, alongside a multi-stage optimized Docker file:

# Terraform Provider AWS declaration
provider "aws" {
  region = "us-east-1"
}

# Auto Scaling Group configuration
resource "aws_autoscaling_group" "app_asg" {
  name_prefix         = "mirahlabs-app-asg-"
  desired_capacity    = 2
  max_size            = 10
  min_size            = 2
  vpc_zone_identifier = ["subnet-12345", "subnet-67890"]

  launch_template {
    id      = aws_launch_template.app_lt.id
    version = "$Latest"
  }

  target_group_arns = [aws_lb_target_group.app_tg.arn]

  tag {
    key                 = "Environment"
    value               = "Production"
    propagate_at_launch = true
  }
}

# Dynamic Scaling Policy based on Target CPU Utilization
resource "aws_autoscaling_policy" "cpu_scaling" {
  name                   = "target-cpu-scaling"
  autoscaling_group_name = aws_autoscaling_group.app_asg.name
  policy_type            = "TargetTrackingScaling"

  target_tracking_configuration {
    predefined_metric_specification {
      predefined_metric_type = "ASGAverageCPUUtilization"
    }
    target_value = 65.0
  }
}

And here is the corresponding multi-stage production Dockerfile to build lightweight, secure images:

# Stage 1: Build dependencies
FROM python:3.11-alpine AS builder
WORKDIR /app
RUN apk add --no-cache gcc musl-dev libffi-dev g++ postgresql-dev
COPY requirements.txt .
RUN pip install --user --no-cache-dir -r requirements.txt

# Stage 2: Final lightweight image
FROM python:3.11-alpine
WORKDIR /app
RUN apk add --no-cache libpq
COPY --from=builder /root/.local /root/.local
COPY . .
ENV PATH=/root/.local/bin:$PATH
EXPOSE 5001
USER 1001
CMD ["gunicorn", "--bind", "0.0.0.0:5001", "run:app"]

Production Trade-offs & Implementation Decisions

Deploying this solution in production environments requires a careful analysis of the trade-offs involved. For instance, focusing purely on consistency (such as ACID compliance) can limit network throughput and horizontal scalability. On the other hand, adopting an eventual consistency model can lead to dirty reads and requires complex conflict resolution strategies in the application layer.

At MirahLabs, our engineering teams balance these architectural constraints by separating critical transaction paths from analytics workloads. We apply message-driven architectures with idempotent consumer systems to guarantee that network failures or retries do not result in double processing or state contamination.

Real-World Benchmarks & Resource Planning

Below is a typical performance comparison profile compiled by our engineering team in staging environments under simulated loads (10k concurrent virtual users):

When capacity planning, we recommend scaling out horizontally using containerized workloads rather than vertically upgrading underlying instance models. This maximizes uptime and provides cost efficiency through dynamic scaling policies.

Security Considerations & Vulnerability Mitigations

No production blueprint is complete without addressing security. Ensure that all data paths utilize encryption in transit (TLS 1.3) and at rest (using AES-256). Furthermore, implement strict Role-Based Access Control (RBAC) to limit operations. For APIs, always enforce rate limits (e.g. using token bucket algorithms in Redis) and run continuous static application security testing (SAST) in your CI pipeline.

How MirahLabs Applies This in Practice

Our experience building high-volume solutions like MirahCare.ai and Ayurveda.ai has taught us that early optimization is often a trap, but ignoring structural security and data design early leads to fatal development blocks. We design all client products from day one to support modular extensions, robust query indexing, and standard schema definitions, ensuring rapid iteration without technical debt growth.