The common tool customers use right now is https://github.com/kubecost. That will allow you to segment cost from E2E within a Kubernetes Cluster.
Unable to segregate cost per client in AKS
Team,
We have a service hosted as a container inside AKS and it is accessible only through my organization. Now there are multiple teams of different verticals are using that API to get some data.
Whenever we make a call from my organization network to the AKS then ingress(nginx) logs only common organisation IP in every request. This is obvious because every organization has NAT registered in such a way.
Due to this issue, I'm unable to trace that which client is consuming how many resources hence difficulty in making a final cost balance sheet per client. The amount of key vault or container registry can still be considered as fixed. In addition, it could have been possible for azure bandwidth as well because I know the size of my payload but problem again arise while tracing the count of calls made by a specific client. So, confusion is there while calculating the chargeback for RAM or CPU consumed in AKS, Standard load balancer and Azure bandwidth.
Can you suggest few approaches using which I can setup the service in such a way that I can calculate the cost per client consuming the resources.
I would really appreciate if you share your valuable suggestion on the same.
Thank you