I hope this email finds you well. I wanted to share some feedback regarding a behavior we encountered while using Skupper in our multi-cluster setup.
We have Skupper configured in two namespaces within the same cluster, and as part of the setup, it appears that Skupper is creating a separate LoadBalancer service for each Skupper router instance in each namespace. While we understand the intention behind having dedicated services for each router, this has led to an unexpected number of LoadBalancer services being created. Given that LoadBalancers are a more limited and often expensive resource in cloud environments, this can be inefficient and costly.
Would it be possible to explore a more optimized approach to this? For instance, there could be an option to share a single LoadBalancer service across multiple router instances within the same cluster, or even an alternative service type configuration that provides more control over LoadBalancer creation.
We appreciate your ongoing work on Skupper and believe this improvement could benefit many users, particularly those working in resource-constrained environments.
Thank you for considering this suggestion, and please feel free to reach out if you'd like more details on our use case or if we can assist further.