Storage Developer Conference

Software memory copies have been the gold standard for applications performing memory data movement or operations in system memory. With new accelerators and memory types enriching the system architecture, accelerator-assisted memory data movement and transformation needs much-needed standardization. SNIA's SDXI (Smart Data Accelerator Interface) TWG is at the forefront of standardizing this and has been working towards a v1.0 since its formation in June 2020. In this talk, SNIA SDXI TWG members will summarize the SNIA SDXI specification’s development journey, its various use cases, and features leading to SNIA SDXI v1.0. Further, this talk will shed some light on future directions for the specification. Learning Objectives: 1) Learn from the experts designing a standard for memory to memory data movement and acceleration; 2) SDXI specification v1.0 internals; 3) SDXI Usecases; 4) Applicability to Accelerators, Persistent Memory, Computational Storage, CXL, and other industry trends.

2022 has been an interesting and challenging year for storage security. The cyber threat landscape has witnessed large numbers of attacks impacting data and increased nation state activities directed at critical infrastructure. The regulatory landscape is undergoing change as well (e.g., EU Directive 2009/125/EC also known as LOT 9) and potentially imposing requirements that necessitate adjustments to security capabilities, controls, and practices to reflect new realities. By the end of 2022 there will be significant changes to security standards and specifications relevant to storage. New technologies could increase the storage security options. Lastly, new practices and deployment strategies could add further data protections. This session concentrates on the new and emerging storage security elements and issues rather than covering storage security from a general perspective. In addition, the session homes in on those aspects that are potentially relevant to developers and architects. Learning Objectives: 1) Understand key threat and regulatory landscape issues that could affect storage development; 2) Identify important security standards and specifications that should be considered in the development of storage products; 3) Recognize the implications and challenges associated with new practices and deployment strategies.

The Key Per IO (KPIO) project is a joint initiative between NVM Express® and the Trusted Computing Group (TCG) Storage Work Group to define a new KPIO Security Subsystem Class (SSC) under TCG Opal SSC for NVMe® class of Storage Devices. Self-Encrypting Drives (SED) perform continuous encryption on user accessible data based on contiguous LBA ranges per namespace. This is done at interface speeds using a small number of keys generated/held in persistent media by the storage device. KPIO will allow large number of encryption keys to be managed and securely downloaded into the NVM subsystem. Encryption of user data then occurs on a per command basis (each command may request to use a different key). This provides a finer granularity of data encryption that enables a granular encryption scheme in order to support use cases: Support of EU - GDPR Support of data erasure when data is spread over many disks, support of data erasure of data that is mixed with other data needing to be preserved (multitenancy), assigning an encryption key to a single sensitive file or host object. The presentation will introduce the architectural differences between traditional SEDs and the KPIO SSC, provide an overview of the proposed TCG KPIO SSC spec and the features in the NVMe commands to allow use of KPIO, and conclude by summarizing the current state of the standardization proposals in NVM Express and the TCG Storage WG. Learning Objectives: 1) Understand how encryption of data at rest protects that data today; 2) Understand how fine grain encryption (KPIO) will be used to protect data at rest in the future; 3) Understand possible use cases for KPIO (multi-tenant use of a common device, EU GDP use cases, others).

NVM Express® (NVMe®) has become synonymous with high-performance storage with widespread adoption in client, cloud, and enterprise applications. The NVMe 2.0 family of specifications, released in June 2021, allow for faster and simpler development of NVMe solutions to support increasingly diverse environments, 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. This presentation provides an overview of the latest NVMe technologies, summarizes the NVMe standards roadmap, and describes new NVMe standardization initiatives. Learning Objectives: 1) Gain an overview of the latest NVMe technologies; 2) Summarize the NVMe standards roadmap; 3) Describe new NVMe standardization initiatives.

Emerging memory technologies have gotten a couple of big boosts over the past few years, one in the form of Intel’s Optane products, and the other from the migration of CMOS logic to nodes that NOR flash, and now SRAM, cannot practically support. Although these appear to be two very different spheres, a lot of the work that has been undertaken to support Intel’s Optane products (also known as 3D XPoint) will lead to improved use of persistent memories on processors of all kinds: “xPUs”. In this presentation we will review emerging memory technologies and their roles in replacing other on-chip memories, the developments through SNIA and other organizations fostered by Optane, but usable in other aspects of computing, the emergence of new Near/Far Memory paradigms that have spawned interface protocols like CXL and OMI, and the emergence of “Chiplets,” and their potential role in the evolution of persistent processor caches. Learning Objectives: 1) New memories and their impact on computing architecture; 2) Near & Far Memory and how it interacts with new persistent memory technologies; 3) How the adoption of chiplets impacts these changes.

Smart Data Accelerator Interface (SDXI) is a proposed standard for a memory to memory data movement and acceleration interface. Software memcpy is the current data movement standard for software implementation due to stable CPU ISA. However, this takes away from application performance and incurs software overhead to provide context isolation. Offload DMA engines and their interface are vendor-specific and not standardized for user-level software. SNIA’s SDXI TWG is tasked with developing and standardizing an extensible, forward-compatible memory to memory data mover and acceleration interface that is independent of actual data mover implementations and underlying I/O interconnect technology. In this panel discussion, experts and representatives of SDXI TWG member companies will talk about their motivations in joining this industry-standard effort. Learning Objectives: 1) Learn from the experts desigining a standard for memory to memory data movement and acceleration; 2) Learn about the use cases of interest to SDXI TWG member companies; 3) Learn about the ecosystem being developed by SDXI member companies for data movers and accelerators.

In this presentation the Co-Chairs of the Computational Storage Technical Working Group (CS TWG) will provide a status update from the work having been done over the last year, including the release of the new Public Review materials around Architecture and APIs. We will update the status of the definition work and address the growing market and adoption of the technology with contributions from the 47+ member organizations participating in the efforts. We will show use cases, customer case studies, and efforts to continue to drove output from the Technical efforts. Learning Objectives: 1) Learn the Latest of the work done by CS TWG; 2) Computational Storage Architectures API Deployment; 3) Customer Case Studies; 4) Market Update and Overview.

For the last eight years Backblaze has collected daily operational data from the hard drives in our data centers. This includes daily SMART statistics from over 250,000 hard drives, and SSDs, totaling nearly two exabytes of storage, and totaling over 200 million data points. We'll use this data to examine the following: - the lifetime failure statistics for all the hard drives we have ever used. - how has temperature effects the failure rate of hard drives. - a comparison the failure rates of helium filled hard drives versus air-filled hard drives. - the SMART stats we use to indicate whether a Hard Drive may fail and if you can use SMART stats to predict hard drive failure. - how SSD failure rates compare to HDD failure rates. The data we use is available for anyone to download and use for their own experimentation and education. Learning Objectives 1) How have the hard drives we have under management performed over time; 2) How you can use SMART stats at scale to determine patterns in hard drive behavior; 3) A look at real word data comparing the failure rates of HDD and SSD devices; 4) How anyone can download and use the data set we have collected.

Current storage technologies include a range of security features and capabilities to allow storage to serve as a last line of defense in an organization’s defense in depth strategy. However, the threat landscape continues to change in negative ways, so new responses are needed. Additionally, the storage technology itself is changing to address the increased capacity and throughput needs of organizations. Technical work in ISO/IEC, IEEE, NVM Express, DMTF, OpenFabric Alliance, Trusted Computing Group (TCG), Open Compute Project (OCP), Storage Networking Industry Association (SNIA), etc. are introducing new storage technologies, specifying the way storage fits into increasingly complex ICT ecosystems, and identifying protection mechanism for data and the systems themselves. Understanding these developments and their interrelationships will be critical for securing storage systems of the future. This session highlights important storage security elements of both current and emerging storage technologies, including encryption, key management, storage sanitization, roots of trust and attestations, secure communications, and support for multitenancy. Like storage, security technologies are also changing, so crypto-agility, protocol changes, and security practices (e.g., zero trust) are explored. Learning Objectives: 1) Understand current storage security guidance and requirements; 2) Identify security aspects of emerging storage technology; 3) Recognize the implications and challenges associated with securing storage.

The SNIA Computational Storage TWG is driving forward with both a CS Architecture specification and a CS API specification. How will these specification affect the growing industry Computational Storage efforts? Learn what is happening in industry organizations to make Computational storage something that you can buy from a number of vendors to move your computation to where your data resides. Hear what is being developed in different organizations to make your data processing faster and allow for scale-out storage solutions to multiply your compute power. Learning Objectives: 1) Learn what is going on in the standards bodies; 2) Learn how the standards can help you utilize Computational Storage; 3) Learn about where Computational Storage is being standardized and how to get involved.

DNA data storage is an attractive option for digital data storage because of its extreme density, durability and eternal relevance. This is especially attractive when contrasted with the exponential growth in world-wide digital data production. In this talk we will present our efforts in building an end-to-end system, from the computational component of encoding and decoding to the molecular biology component of random access, sequencing and fluidics automation. We will also discuss some early efforts in building a hybrid electronic/molecular computer system that can offer more than just data storage, for example, image similarity search. Learning Objectives: 1) Describe the properties of synthetic DNA as a digital data storage medium; 2) Describe how an end-to-end DNA data storage system works; 3) Describe how to perform certain types of computation with synthetic DNA molecules.

The SNIA Swordfish specification has expanded to include full NVMe and NVMe-oF enablement and alignment across DMTF, NVMe, and SNIA for NVMe and NVMe-oF use cases. This presentation will provide an overview of the most recent work adding detailed implementation requirements for specific configurations, ensuring NVMe and NVMe-oF environments can be represented entirely in Swordfish and Redfish environments. Learning Objectives: 1) Describe how the NVMe and NVMe-oF environments can be managed in Swordfish and Redfish; 2) Provide an overview of current work in progress to extend NVMe and NVMe-oF manageability in Swordfish; 3) Describe the updated guidance for implementations in Swordfish profiles and documentation.

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. The PCIe 6.0 specification – targeted for final release in 2021 – will deliver 64 GT/s data rate (256 GB/s via x16 configuration), while maintaining backward compatibility with previous generations. In this session, attendees will learn the nuts and bolts of PCIe 6.0 architecture and how it will enable high-performance networking. Some key features of the upcoming specification include PAM4 encoding, low-latency Forward Error Correction (FEC), and backward compatibility with all previous generations of PCIe technology. This presentation will also highlight PCIe 6.0 technology use cases and the heterogenous computing applications that will be accelerated by PCIe 6.0 technology, including artificial intelligence, machine learning and deep learning. Finally, attendees will receive an update on the release timeline of the PCIe 6.0 specification later this year and rollout of the interoperability and compliance program. Learning Objectives: 1) Attendees will learn the nuts and bolts of PCIe 6.0 architecture and how it will enable high-performance networking; 2) Key features of the upcoming PCIe 6.0 specification; 3) PCIe 6.0 technololgy use cases and the heterogenous computing applications that will be accelerated by PCIe 6.0.

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. This knowledge is critical when deciding how to build, deploy, operate, and evaluate the performance of a SAN as well as understanding end to end hardware and software implementation tradeoffs. Agenda SAN Transport History and Overview Protocol History Protocol Comparisons NVMe/FC Deep Dive NVMe/RoCE Deep Dive NVMe/TCP Deep Dive Networking Configurations and Topologies for NVMe-oF QoS, Flow Control and congestion L2 local vs L3 routed vs Overlay Offload Processing Considerations and Comparisons Learning Objectives: 1) Cross Comparison of SAN transports; 2) First principles behavior for NVMe-oF transports; 3) Practical networking considerations for deploying NVMe-oF; 4) Implications of NVMe-oF transports for data flow and packet processing.

Traditional Storage Node consists of Compute, Networking and Storage elements. In this case, the entire node is a single failure domain and as such both data and meta data are maintained in storage. Emergence of CXL allows us to re-think the traditional storage node architecture. In future, the storage (behind CXL IO) and metadata memory (behind a CXL memory) can be disaggregated locally or across a bunch of storage nodes to improve the availability of the data. Further, memory persistence can be achieved at a granular level using CXL memory devices. Future extensions to CXL with fabric like attributes have potential to further extend the data replication capabilities of the storage platform. In this talk, we will discuss the various platform architecture options that are emerging for the storage node and how they can change the face of traditional storage node organization. Learning Objectives: 1) Illustrate that the current architecture assumptions for Storage node need to be revisited; 2) Explore options for new storage node architecture with CXL; 3) Explore architecture options for storage node with a fabric extension (beyond today's CXL); 4) Encourage partnership with industry to work together on storage node innovation; 5) Explore storage infrastructure disaggregation and its value to future of storage.

Adopting new memory technologies such as Persistent Memory and CXL-attached Memory is a challenge for software. While libraries and frameworks (such as PMDK) can help developers build new software around emerging technology, legacy software faces a more severe challenge. At IBM Research Almaden we are exploring a new approach to managing heterogeneous memory in the context of Python. Our solution, PyMM, focuses on ease-of-use and is aimed primarily at the data science community. This talk will outline PyMM and discuss how it is being used to manage Intel Optane persistent memory. We will review the PyMM programming abstractions and some early data science use-cases. PyMM is currently an early research prototype with open source availability. Learning Objectives: 1) Understand the emergence of heterogeneous memories (e.g., Optane, CXL-attached); 2) Understand the challenges facing integration of legacy s/w with new memory technology; 3) Introduction and demonstration of PyMM; 4) Outline of some existing use cases.

Storage interfaces have evolved more in the past 3 years than in the previous 20 years. In Linux, we see this happening at two different layers: (i) the user- / kernel-space I/O interface, where io_uring is bringing a low-weight, scalable I/O path; and (ii) and the host/device protocol interface, where key-values and zoned block devices are starting to emerge. Applications that want to leverage these new interfaces have to at least change their storage backends. This presents a challenge for early technology adopters, as the mature part of the Linux I/O stack (i.e., the block layer I/O path) might not implement all the needed functionality. While alternatives such as SDPK tend to be available more rapidly, the in-kernel I/O path presents a limitation. In this talk, we will present how we are enabling an asynchronous I/O path for applications to use NVMe devices in passthru mode. We will speak to the upstream efforts to make this path available in Linux. More specifically, we will (i) detail the changes in the mainline Linux kernel, and (ii) we will show how we are using xNVMe to enable this new I/O path transparently to applications. In the process, we will provide a performance evaluation to discuss the trade-offs between the different I/O paths in Linux, including block I/O io_uring, passthru io_uring, and SPDK. Learning Objectives: 1) Understand the value of I/O passthru; 2) Understand the changes merged into the Linux kernel to support NVMe I/O Pasthru; 3) Understand how to leverage this new I/O path without application changes through xNVMe.

NVMe/TCP has the potential to provide significant benefits in application environments ranging from the Edge to Data Center. However, to fully unlock its potential, we first need to overcome NVMe over Fabrics' discovery problem. This discovery problem, specific to IP based fabrics, can result in the need for the Host admin to configure each Host to access the appropriate NVM subsystems. In addition, any time an NVM Subsystem is added or removed, the Host admin needs to update the impacted hosts. This process of explicitly updating the Host any time a change is made does not scale when more than a few Host and NVM subsystem interfaces are being used. Also, due to the de-centralized nature of this process, it also adds complexity when trying to use NVMe-oF in environments that require high-degrees of automation. For these and other reasons, Dell Technologies, along with several other companies, have been collaborating on innovations that enable an NVMe-oF IP Based fabric to be centrally managed. These innovations, being tracked under nvme.org’s Technical Proposals TP-8009 and TP-8010, enable administrators to set a policy that defines the relationships between Hosts and the NVM subsystems they need to access. These policies are then used by a Centralized Discovery Controller to allow each Host to automatically discover and connect to only the appropriate NVM subsystems and nothing else. Learning Objectives: 1) Explain NVMe-oF’s discovery problem; 2) Review the network topologies that can support the automated Discovery of NVMe-oF Discovery Controllers; 3) Explore the differences between a FC SAN and an IP based SAN used to transport NVMe-oF/TCP; 4) Understand the proposed discovery process through in-depth review of the discovery protocol.

Malware, short for malicious software, is a blanket term for viruses, worms, trojans and other harmful software that attackers use to damage, destroy, and gain access to sensitive information; software is identified as malware based on its intended use, rather than a particular technique or technology used to build it. Ransomware is a blended malware attack that uses a variety of methods to target the victim’s data and then requires the victim to pay a ransom (usually in crypto currency) to the attacker to regain access to the data upon payment (with no guarantees). However, the landscape is changing, and ransomware is no longer just about a financial ransom. Attacks are now being aimed at the infrastructure and undermining public confidence, witness recent headlines regarding incidents affecting police informant databases and oil pipeline sensors. There is also the recent US Treasury guideline to businesses advising them not to pay the ransom. What can we realistically do to prevent such attacks, or do we simply surrender and accept we will lose our data and that the insurance payout will cover any loss? There is increasing evidence that the insurance companies are unwilling to meet those claims, so the situation is perilous as the criminals always appear one step ahead. As a starting point, everyone needs to start assuming they will be attacked at some stage – therefore prevention and mitigation strategies should be based on that assumption. This session outlines the current threats, the scale of the problem, and examines the technology responses currently available as countermeasures. What can be done to prevent an attack? What works and what doesn’t? What should storage developers be thinking about when developing products that need to be more resilient to attack? Learning Objectives: 1) Current ransomware trends and scale; 2) Effectiveness of current data protection technology; 3) What other defensive measures should be considered.

Blockchain has revolutionized decentralized finance, and with smart-contracts has enabled the world of Non-Fungible Tokens, set to revolutionize industries such as art, collectibles and gaming. Blockchains, at the very core, are distributed chained hashes. They can be leveraged to store information in a decentralized, secure, encrypted, durable and available format. However, some of the challenges in Blockchain stem from the bloat of storage. Since each participating node will keep a copy of the entire chain, the same data gets replicated on each node, and even a 5MB file stored on the chain can exhaust systems. Several techniques have been used by different implementations to allow Blockchains for distributed storage of data. The advantages compared to cloud storage would be the decentralized nature of storage, the security provided by encrypting content, and the costs. In this session, we will discuss how different blockchain implementations such as Storj, InterPlanetary File System, YottaChain, and ILCOIN have solved the problem of storing data on the chain, but avoiding bloat. Most of these solutions store the data off-chain and store the transactions metadata on the blockchain itself. IPFS & Storj for example, uses content-addressing to uniquely identify each file in a global namespace connecting all the computing devices. The incoming file is encrypted, and split into smaller chunks, and each participating node will store a chunk with zero-knowledge of the other chunks. ILCOIN relies on RIFT protocol to enable two level blockchains, one for the standard blocks, and the other for the mini-blocks that comprise the transactions and which are not mined, but generated by the system. Yottachain uses deduplication after encrypting content, which is not generally the way data storage is designed for cloud, to reduce the footprint of data on the chain. We will discuss the tradeoffs of these solutions and how they aim to disrupt cloud storage. We will compare the benefits provided in terms of security, scalability and costs, and how organizations such as Netflix, Box, Dropbox can benefit from leveraging these technologies. Learning Objectives: 1) Learn about the Blockchains and how they are designed to store small amounts of information; 2) Learn about different blockchain projects such as IPFS, Yottachain, Storj, ILCOIN and their implementations; 3) How blockchain based storage solutions can provide better benefits to existing cloud storage; 4) Impact of leveraging blockchain storage on companies such as Netflix, Box, Dropbox.