# Kubernetes service with true round-robin

**URL:** <https://discuss.kubernetes.io/t/kubernetes-service-with-true-round-robin/23348>\
**Category:** General Discussions\
**Created:** [March 7, 2023, 1:05pm UTC](https://discuss.kubernetes.io/t/kubernetes-service-with-true-round-robin/23348 "2023-03-07T13:05:53Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![ahmadin](https://sea2.discourse-cdn.com/flex016/user_avatar/discuss.kubernetes.io/ahmadin/32/11989_2.png) [@ahmadin](https://discuss.kubernetes.io/u/ahmadin)\
**Post date:** [March 7, 2023, 1:05pm UTC](https://discuss.kubernetes.io/t/kubernetes-service-with-true-round-robin/23348/1 "2023-03-07T13:05:53Z")

</div>

Asking for help? Comment out what you need so we can get more information to help you!

### Cluster information:

Kubernetes version: 1.24  
Cloud being used: (put bare-metal if not on a public cloud)  
Installation method:  
Host OS:  
CNI and version:  
CRI and version:

You can format your yaml by highlighting it and pressing Ctrl-Shift-C, it will make your output easier to read.

---

<div class="post-metadata">

**Author:** ![ahmadin](https://sea2.discourse-cdn.com/flex016/user_avatar/discuss.kubernetes.io/ahmadin/32/11989_2.png) [@ahmadin](https://discuss.kubernetes.io/u/ahmadin)\
**Post date:** [March 7, 2023, 1:13pm UTC](https://discuss.kubernetes.io/t/kubernetes-service-with-true-round-robin/23348/2 "2023-03-07T13:13:23Z")

</div>

I have a requirement to use round-robin for a service.  
I have a rest api served by 4 pods. This service is using kafka topic that has only 4 partitions. I can only have 4 pods and each pod is assign to a partition (using consumer group).  
The rest api have many consumers (more than 4). Unfortunately, with this setup, it seem that the service doesn’t use round-robin to distribute the calls to the service. It always go to the first pod until a certain load (it seems that the distribution to be 50% on the first pod, 30% on second… till 0% to the 4th pod). The problem for us is that messages in kafka are distributed to all partitions (round-robin, I think) and causing some of messages from some partitions are not delivered to the service consumer.  
Is there a way to force the service to use round-robin to distribute load to the pods?