This podcast discusses distributed cloud storage with Enrico Signoretti, the vice-president of product and partnerships at Cubbit. He explains that distributed cloud storage separates the control plane from data, allowing for data retention in multiple locations, both on-site and in multiple clouds. Signoretti emphasizes that distributed cloud is important for companies as they transition from traditional models to hybrid and multicloud environments.
In a distributed cloud model, the control plane is kept by the service provider while resources can be deployed anywhere, including public cloud environments or on-premise datacenters. This distribution of resources can also be geo-distributed, meaning that data can be stored in different geographical locations. This concept is important as it removes obstacles in working with multicloud environments.
Signoretti discusses the advantages of distributed cloud over public cloud and on-site datacenters. Distributed cloud offers control, data sovereignty, and data independence, allowing organizations to choose where they store their data and maintain control at all levels of data management. This level of control is particularly valuable in regions with strict data regulations where giving data to hyperscalers is not feasible. Distributed cloud provides a middle ground between on-premise and public cloud, offering the benefits of both while allowing organizations to retain control.
When considering workloads for distributed cloud, Signoretti suggests that low latency and high-performance workloads that require close proximity between the CPU and storage are not ideal for distributed cloud. However, other workloads such as backup, disaster recovery, collaboration, big data lakes, and storing data for AI and ML applications are well-suited for distributed cloud. Although there may be slightly higher latency compared to on-premise or public cloud environments, the throughput remains good for most use cases.
