Many large-scale vendors are still grappling with migrating to Kubernetes, particularly when they are operating a large number of Kubernetes clusters, often across different architectures and regions.
In this episode of TFiR Let’s Talk recorded at the KubeCon in Detroit, Swapnil Bhartiya catches up with Julian Fischer, CEO and Founder of anynines, to discuss how the platform is helping organizations running Postgres across multiple Kubernetes clusters over multiple infrastructure regions. He goes into the key challenges of distributed databases and how anynines is helping tackle these challenges.
Key highlights from this video interview:
- The next-generation application runtime for Cloud Foundry is on its way. Fischer tells us it is to a maximum degree Kubernetes native and the next step would be to enable it to maintain similar operational efficiency, particularly for large Cloud Foundry environments, to migrate.
- Fischer explains that some large-scale vendors are struggling to operate a large number of Kubernetes clusters. He discusses how the BOSH-based virtual machine automation anynines data services enable organizations to efficiently manage data the same way with Cloud Foundry runtime using Bosh. He tells us how this works in Kubernetes and why there is still a need for more technologies that allow for operational efficiency and sets of tools that help address the challenges of data management.
- anynines is testing several working hypotheses, including using the platform across on-premise and public infrastructures, across multiple Kubernetes distros.
- The company is building data service automation natively on Kubernetes in the Kubernetes section of its platform. Fischer explains that it will be executed under the same license as the anynines data services.
- Beta version 3 of the Postgres operator is now available and they are actively looking for startup customers to give them feedback about it.
- One of the main challenges of dealing with distributed databases is how to test network issues. To address this, anynines has built a testing framework that allows users to simulate network partitions. Fischer explains how this is helping improve predictability.
- At KubeCon, Fischer talked about how to stretch a Kubernetes cluster, for instance, if you have geographic regions and would like the database to span across those regions. He explains that the Cilium network plugin can help people achieve this. They are looking forward to having customers with demand for inter-cluster Postgres instances.
The summary of the show is written by Emily Nicholls.