How to address IP Exhaustion for large scale workload running ECS with Fargate Launch Type

Question

Hi guys,
I'm running a large-scale workload on the ECS Fargate launch type in a VPC CIDR (10.0.0.0/16). The Task that runs Fargate are located in a Private subnet with subnet mask /22. However, these tasks have scaled significantly recently and we are running out of IP addresses. I have read an article about addressing IP exhaustion on ECS with EC2 launch type that recommended the use of ENI Trunking to a subnet in a secondary CIDR in VPC. However, I'm not sure this will work for the Fargate Launch type.
If you know if is it possible to use the ENI Trunking with secondary CIDR in VPC for Fargate, please let me know.

Thanks,
Steven

Answer

ENI trucking is only supported with the EC2 deployment model.

You could add additional subnets to your ECS configuration. You don’t mention you are using multiple AZs.

If your out of addresses You could add a 2nd IP CIDR range to your VPC and move your ECS to the new range with a bigger network.

How to address IP Exhaustion for large scale workload running ECS with Fargate Launch Type

Relevant content