1 Answer
- Newest
- Most votes
- Most comments
0
Hi, effectively if you change the target to instance instead of ip, then every node will register as target in the targetgroup, so if you have <100 nodes in the cluster this can be a good solution.
You can find more details in the EKS best practice guide : https://aws.github.io/aws-eks-best-practices/networking/loadbalancing/loadbalancing/#use-ip-target-type-load-balancers
If you have > 100 nodes, this will not solve the problem, and instead you could also consider using an internal ingress like nginx, and only register the nginx service pods in the ALB/NLB, and then let the ingress do the routing inside the cluster
answered a year ago
Relevant content
- asked 2 months ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago