Hi all, We are ready to start working with Airbyte within the datawarehouse deployments of our clientele. As we are servicing multiple clients on different cloud projects (mainly GCP) I was wondering what the best setup would be to use Airbyte in a scalable but secure way. For some - but not all - projects we are also looking towards using Airbyte in combination with Airflow in order to schedule successive tasks. We run Airflow using kubernetes (similar to https://airflow.apache.org/docs/apache-airflow/stable/executor/kubernetes.html#fault-tolerance). Currently we are considering multiple options and I wanted to reach out to you guys to hear you opinion about this case as we could not find any blogs/documents about a similar setup. The following options are considered: 1. Airbyte deployment on a separate VM for each client project (hard to scale, but secure as only people with the right permissions to the project can access the VM) 2. One general Airbyte VM where all the connections are set up (scalable but prone to reach limitations and not very secure as the whole team needs to be able to access this VM for setting up connections) 3. Running Airbyte on Kubernetes with a separate server on the cluster for each client project (scalable but not sure if this will be secure as I don't know how we will access those servers without serving them to the internet as RBAC is not yet supported) 4. Other option(s) I would really value your opinion on this as I am not a cloud engineer by heart and would like to make a well-considered decision. Thanks!

Hi @Guus van Heijningen, I'd suggest you to leverage Kubernetes and run multiple Airbyte deployment in the same cluster, on different namespaces. The use of different K8S ingresses for each cluster + IP filtering can help you to isolate deployments and restrict access on them.