For many edge computing projects, HA is a required capability. Mission-critical edge workloads demand resiliency and fault tolerance.But the cost, space and power needed for a conventional 3-node+ Kubernetes architecture, using the quorum-based etcd data store, is often prohibitive at edge scale.A 2-node HA approach would produce dramatic savings in hardware and other costs. But until now, it has only been possible with proprietary technologies, an external data store, or a third ‘witness’. These approaches all have drawbacks.In this webinar we introduce a new 2-node HA architecture for edge Kubernetes, based on OSS technologies. It delivers the fault tolerance you need, without the tradeoffs of these other approaches, or the risk of a ‘split brain’ scenario.We’ll explore:The common use cases for edge HA and the implications of a 3-node architectureThe technology landscape and some of the alternative approaches available for delivering HAThe rationale for our 2-node design and the OSS projects, such as K3s, that it incorporatesConsiderations and best practices for architecting your edge deployment for HAWe’ll show you a demo of this 2-node architecture in action and show you how you can try it for yourself.For more details, check out our recent blog, or come and speak to us at Edge Day at KubeCon in Chicago!
Recorded: Wednesday November 1, 2023