Idle Load Balancer Cost

How to detect idle load balancer cost

The strongest signals are no recent traffic, no registered targets, or both at the same time on a load balancer that is still deployed.

Cloud Waste Hunter reviews application, network, and classic load balancers over the last 14 days. It looks for either of these signals:

• no recent traffic in the relevant CloudWatch metric
• no registered targets behind the load balancer

That makes the detector useful for both obviously abandoned load balancers and half-removed environments where the ELB shell is still standing after the workload moved on.

Inventory the load balancers first:

aws elbv2 describe-load-balancers
aws elb describe-load-balancers

Then check whether traffic and targets still exist:

aws cloudwatch get-metric-statistics \
  --namespace AWS/ApplicationELB \
  --metric-name RequestCount \
  --dimensions Name=LoadBalancer,Value=app/my-alb/1234567890abcdef \
  --start-time 2026-03-09T00:00:00Z \
  --end-time 2026-03-23T00:00:00Z \
  --period 86400 \
  --statistics Sum

aws elbv2 describe-target-health \
  --target-group-arn arn:aws:elasticloadbalancing:us-east-1:123456789012:targetgroup/app/1234567890abcdef

The best cleanup candidates are load balancers with no meaningful traffic, no registered targets, weak ownership, and no DNS or routing reason to keep them.

Caveats and false-positive scenarios

This detector does not know whether a quiet load balancer is still strategically important. Disaster recovery endpoints, internal services with sporadic traffic, or cutover infrastructure can all look idle while still being intentional.

That makes Unassociated Elastic IPs a good companion page. If both the front-end IP allocations and the load balancer look stale, the environment cleanup case is much stronger.

How to fix idle load balancer cost

Treat load balancer cleanup as a small network teardown, not just a delete command:

1. Confirm no clients, DNS records, or route rules still rely on it.
2. Remove or migrate any remaining listeners and target groups.
3. Delete the load balancer after rollback expectations are clear.

aws elbv2 delete-load-balancer \
  --load-balancer-arn arn:aws:elasticloadbalancing:us-east-1:123456789012:loadbalancer/app/app-lb/1234567890abcdef

How this differs from nearby detectors

Unassociated Elastic IPs is about reserved public IP allocations with no active attachment. Idle RDS instances is about database footprint that no longer shows meaningful workload activity. This detector is specifically about ELB resources whose traffic or target picture no longer justifies keeping them online.

How Cloud Waste Hunter helps

Cloud Waste Hunter can surface idle load balancers with type, region, traffic evidence, target-count evidence, and estimated savings so teams can fold ELB cleanup into broader AWS environment rationalization. For related reviews, continue into the AWS Idle and Underused Resources guide.