# Unconsolidated Application Load Balancers per Service

Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/unconsolidated-application-load-balancers-per-service

A single Application Load Balancer can route traffic for many applications using listener rules on host names and URL paths, forwarding each to its own...

By: PointFive

Updated: 2026-09-28

[Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) 

The short version

A single Application Load Balancer can route traffic for many applications using listener rules on host names and URL paths, forwarding each to its own target group.

PointFive Research

Cloud cost research at PointFive

AWS service

[AWS ELB](https://www.pointfive.co/efficiency-hub/cloud-services/aws-elb)

Category

[Networking](https://www.pointfive.co/efficiency-hub/service-category/networking)

Reference

CER-0359

Type

Inefficient Architecture

## Explanation

Why the waste happens and who it affects.

In practice many organizations give every microservice, environment or team its own ALB. On Amazon EKS the AWS Load Balancer Controller creates a separate ALB for each Ingress unless Ingresses are grouped, which AWS notes can result in more load balancers than necessary and recommends sharing ALBs across Ingresses to reduce costs. ECS services and infrastructure-as-code modules often follow the same one-service-one-ALB pattern.

Every ALB carries its own hourly charge for as long as it exists, plus LCU charges, and each internet-facing ALB also consumes public IPv4 addresses that are billed separately. For low-traffic services the fixed hourly charge dominates, so dozens of lightly used ALBs cost much more than a few shared ones handling the same requests.

## Billing model

The pricing dimensions that drive this cost.

ALB-hour

Each Application Load Balancer is billed for every hour or partial hour it runs; AWS lists $0.0225 per ALB-hour in US East (N. Virginia)

LCU-hour

Billed on the highest of new connections, active connections, processed bytes and rule evaluations; the first 10 processed rules per request are free

Public IPv4 addresses

Internet-facing load balancers incur standard public IPv4 address charges for every address they consume

## How to detect

5 checks to find it in your estate.

- Inventory ALBs per account, Region and VPC and compare their counts with the number of distinct applications; many ALBs with low RequestCount and ConsumedLCUs in CloudWatch are consolidation candidates

- Flag ALBs with a single listener rule or a single target group, which typically front one small service

- In EKS clusters, list Ingresses with ingressClassName alb (or the legacy class annotation) that lack the alb.ingress.kubernetes.io/group.name annotation; each of these gets its own ALB

- In ECS, list services that each reference a target group on a different ALB even though they share a domain, VPC and security posture

- Count public IPv4 addresses attributed to load balancers in the VPC IP Address Manager or Cost and Usage Report to see the address charges each additional internet-facing ALB adds

## How to fix

5 ways to remove the waste.

- Consolidate compatible low-traffic services onto shared ALBs using host-header and path-pattern listener rules, one target group per service

- In EKS, add the same alb.ingress.kubernetes.io/group.name annotation to Ingresses that should share an ALB and use group.order to control rule precedence; only do this when everyone with permission to create or modify Ingresses is within the same trust boundary, because any member can add or override rules on the shared ALB

- Stay within ALB quotas when consolidating, such as 100 rules per ALB (adjustable), 100 target groups per ALB and 25 certificates per ALB (adjustable), and account for rule-evaluation LCUs when a shared ALB has many rules

- Keep separate ALBs where isolation genuinely matters, such as different security or WAF policies, compliance boundaries, blast-radius limits, or internal versus internet-facing exposure

- Delete the redundant ALBs and their listeners after DNS and target traffic have moved to the shared load balancers

## Documentation

Vendor references for pricing and configuration.

- [Elastic Load Balancing pricing  aws.amazon.com](https://aws.amazon.com/elasticloadbalancing/pricing/)

- [Condition types for listener rules  docs.aws.amazon.com](https://docs.aws.amazon.com/elasticloadbalancing/latest/application/rule-condition-types.html)

- [Quotas for your Application Load Balancers  docs.aws.amazon.com](https://docs.aws.amazon.com/elasticloadbalancing/latest/application/load-balancer-limits.html)

- [A deeper look at Ingress Sharing and Target Group Binding in AWS Load Balancer Controller  aws.amazon.com](https://aws.amazon.com/blogs/containers/a-deeper-look-at-ingress-sharing-and-target-group-binding-in-aws-load-balancer-controller/)

- [Ingress annotations - AWS Load Balancer Controller  kubernetes-sigs.github.io](https://kubernetes-sigs.github.io/aws-load-balancer-controller/latest/guide/ingress/annotations/)

## Related inefficiencies

[Browse the library](https://www.pointfive.co/efficiency-hub)

- AWS ELB  CER-0187

### [Elastic Load Balancer with Only One EC2 Instance](https://www.pointfive.co/efficiency-hub/inefficiencies/elastic-load-balancer-with-only-one-ec2-instance)

An ELB with only one registered EC2 instance does not achieve its core purpose-distributing traffic across multiple backends. In this configuration, the ELB adds complexity and cost without improving availability, scalability, or fault...

Networking

- AWS ELB  CER-0059

### [Inactive Classic Load Balancer (CLB)](https://www.pointfive.co/efficiency-hub/inefficiencies/inactive-classic-load-balancer-clb)

Classic Load Balancers that no longer serve active workloads will persist if they are not properly decommissioned. This often happens after application migrations, architecture changes, or testing activities. Even if no connections or...

Networking

- AWS ELB  CER-0006

### [Inactive Network Load Balancer (NLB)](https://www.pointfive.co/efficiency-hub/inefficiencies/inactive-network-load-balancer-nlb)

Network Load Balancers that are no longer needed often persist after architecture changes, service decommissioning, or migration projects. When no active TCP connections or traffic flow through the NLB, it still generates hourly...

Networking

---
Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.

