In this presentation, I will delve into the world of NVMe RAID and present a comprehensive analysis and comparison of various NVMe RAID engines. Through comparative testing and performance analysis using state-of-the-art performance analysis tools, I have identified bottlenecks in these engines when handling different workloads. During the presentation, I will provide valuable insights and comparisons, showcasing significant figures that highlight the performance differences among the tested NVMe RAID solutions.

In recent years, zoned storage has become pervasive across the storage software ecosystem, including file systems, cloud storage, and end-to-end application integrations. Zoned storage overcomes the drawbacks of write amplification by enabling the host to collaborate with the storage device when submitting writes. The ecosystem seamlessly integrates support for shingled magnetic drives (SMR HDDs), SSDs with Zoned Namespaces (ZNS SSDs), and UFS-enabled mobile devices through a single storage abstraction.

The demand for data storage continues to grow exponentially with the overall data storage temperature cooling down with most data becoming cold after one month and subsequently infrequently accessed. Given the evolving storage pyramid and the growth of the overall storage ecosystem, especially cold storage, demand for significant capacity, longevity and sustainability scaling is increasing.

Hyperscale deployments are using Serial Attached SCSI (SAS) today, to solve capacity and cost challenges within their large-scale storage infrastructure. Specific features within SAS address these problems and offer developers real solutions. This presentation will discuss how SAS addresses the scaling, reliability, cost and flexibility requirements of modern hyperscale architectures and show how this maps to Meta's Grand Canyon architecture.

With the complexity of applications increasing every day, the workloads generated by these applications are complicated and hard to replicate in test environments. We propose an efficient method to synthesize a close approximation of these application workloads based on analyzing the historic autosupport data from field using an iterative mechanism and also a method to store and replay these workloads in the test environment for achieving the goals of customer driven testing.