We are all used to thinking that from a performance perspective distributed storage symmetrical systems perform best because of the weakest link in the chain effect. This presentation discusses situations in which asymmetrical implementation reduces the cost of implementation and improves the performance. This happens because of the public cloud resource variety and price structure. The presentation discusses the model and the details for cheaper and more performant asymmetrical Ceph deployment

The advent of cloud and the everything as a service (XaaS) model requires storage developers to rethink how their products are consumed. Organizations are looking to develop infrastructure and processes that are agnostic from any one cloud vendor, including their own on-premises datacenters. Container orchestrators (COs) like Kubernetes enable this ideal by allowing entire application deployments to be packaged up (in containers and manifests) and moved from environment to environment with relative ease.

This is a two-part detailed technology presentation on LTO, erasure codes, and archival storage. In the first part, we will cover core LTO technology by reviewing areal density roadmaps, physical and logical LTO formats, and data durability for long-term archival applications including environmental requirements. In the second part, we will discuss new two-dimensional erasure-coded tape for durable and available long-term data storage where random access is used. Tape has recently reemerged as the lowest cost medium for different levels of cold data in archival data storage.

Data has become the new source code and data storage and I/O are becoming the major inhibitor for deriving actionable intelligence as well as proving data management capabilities for managing data - DataOps, is becoming paramount for the success of AI projects, particularly in Capital markets. This session will discuss the trends, solutions and benchmarks involved for successful AI imprementation for historic and realtime datasets.

Yes, it really does say Windows and SPDK in the same sentence! The Storage Performance Development Kit (SPDK) is a well-regarded set of tools and libraries for writing high performance user mode storage applications on Linux and FreeBSD. However, in a typical Data Centre, a significant percentage of the servers will be running Microsoft Windows where the options for NVMe support are more limited. This talk looks at what is involved in making SPDK run natively on Windows.

The NVM Express® (NVMe®) family of specifications, released in June 2021, allow for faster and simpler development of NVMe solutions to support the increasingly diverse NVMe device environment, now including Hard Disk Drives (HDDs). The extensibility of the specifications encourages the development of independent command sets like Zoned Namespaces (ZNS) and Key Value (KV) while enabling support for the various underlying transport protocols common to NVMe and NVMe over Fabrics (NVMe-oF™) technologies.

Over the past nearly three decades, PCI-SIG® has delivered a succession of industry-leading specifications that remain ahead of the curve of the increasing demand for a high-bandwidth, low-latency interconnect for compute-intensive systems in diverse market segments, including data centers, PCs and automotive applications. Each new PCI Express® (PCIe®) specification consistently delivers enhanced performance, unprecedented speeds, and low latency – doubling the data rate over previous generations.

This talk will discuss the latest trends in the growth of emerging non-volatile memories and look beyond at the emergence of new computer architectures which will use non-volatile memories for near memory and shared far memory. Emerging non-volatile memory technologies (such as 3D XPoint, MRAM and RRAM) are now available in the data center and in the next generation of AI-based IoT devices. Major foundries are making SoCs with MRAM and RRAM replacing NOR flash and SRAM. Non-volatile memory will play a big role in the data center, at the edge and endpoints.

Continuous Integration/Continuous Delivery (CI/CD) platforms such as Jenkins, Bamboo, CircleCI and the like dramatically speed deployments by allowing administrators to automate virtually every process step. Automating the data layer is the final frontier. Today, hours are wasted manipulating datasets in the CI/CD pipeline. Allowing Kubernetes to orchestrate the creation, movement, reset, and replication of data eliminates dozens of wasted hours from each deployment, accelerating time to market by 500X or more.

Abstract Block storage access across Storage Area Networks (SANs) have an interesting protocol and transport history. The NVMe-oF transport family provides storage administrators with the most efficient and streamlined protocols so far leading to more efficient data transfers and better SAN deployments. In this session we will explore some of the protocol history to set the context for a deep dive into NVMe/TCP, NVMe/RoCE, and NVMe/FC. We will then examine network configurations, network topology, QoS settings, and offload processing considerations.