Is there a way to implement kafka-broker autoscaling with Strimzi Operator. #10562
Replies: 3 comments 2 replies
-
If you use KafkaNodePools, you can scale tham as any other Kubernetes resource. Then you need to do rebalancing to move that data - that will be automated one day, but currently you have to do it yourself. But Kafka will never scale-up and down for few minutes - it will never have that kind of elasticity (or at least not anytime soon). So there migth be situations where it adds some value, but also many where it would not. |
Beta Was this translation helpful? Give feedback.
-
Thank you @scholzj ! |
Beta Was this translation helpful? Give feedback.
-
Thank you @ppatierno ! https://strimzi.io/blog/2023/08/14/kafka-node-pools-introduction/ |
Beta Was this translation helpful? Give feedback.
-
Hello Team,
Could anyone assist with implementing Kafka broker autoscaling using the Strimzi Operator? Specifically, I'm looking for ways to enable autoscaling for a Kafka cluster managed by Strimzi, where the number of replicas can dynamically adjust between 5 and 10 pods based on consumption.
I am using Strimzi within an AKS environment and am familiar with autoscaling mechanisms like Cluster Autoscaler, HPA, and VPA. However, I couldn't find any specific information on autoscaling Kafka brokers with Strimzi.
Is there a way to achieve this with the Strimzi Operator? If not, is there any plan to include this feature in the roadmap, or is it considered impractical for Kafka?
Thank you !
Beta Was this translation helpful? Give feedback.
All reactions