Regatta Storage was created to fill a gap in the cloud storage market that was not being met by block or file storage. The system automatically expands with the user's application and synchronizes with S3 using native file formats. When data is not actively being used, it is removed from the Regatta cache, meaning users only pay for the backing S3 storage. The company is also developing a custom file protocol to achieve local-like performance for small-file workloads and Lustre-like scale-out performance for distributed data jobs.
Key takeaways:
- Regatta Storage is a new cloud file system that provides unlimited pay-as-you-go capacity, local-like performance, and automatic synchronization to S3-compatible storage.
- Regatta is designed to automatically expand with your application and only charges for the backing S3 storage when data isn't actively being used.
- Regatta is developing a custom file protocol to achieve local-like performance for small-file workloads and Lustre-like scale-out performance for distributed data jobs.
- Regatta Storage is being used by teams to build serverless Jupyter notebook servers, as a distributed caching layer on top of S3, and to replace thin-provisioned Ceph boot volumes for significant savings.