Storage Developer Conference

While the “perfect” workload to test an array for a customer is the workload or combination that will run against the array in production, enterprises and vendors benefit greatly from a workload approach that more accurately models a production workload to be executed against an array. Enterprises benefit by seeing the characteristics of their own workload or combination, while vendors benefit by being able to scale emulated workloads to array limits and better able to deal with performance issues that occur only after an array is released.

Spatial locality, temporal locality emulating application “burstiness” ensure no array is unrealistically penalized nor given a “free pass” because a test does not reflect the application behavior that will be seen in production.

This is a non-trivial exercise. New approaches use storage controller performance measurement or wire data to derive application workload models that make it possible to test how an application or application combination affects array performance.

Persistent memory NVDIMMs create security challenges, some of which are familiar while others are new. The NVM Programming TWG has been working on a security threat model for Persistent Memory that exposes gaps where solutions need to evolve beyond those that met the needs of hard disks and SSD’s. This session describes the current work in progress of the TWG that will appear in a future white paper.

Learning Objectives: 1)Learn what how persistent memory security solutions may differ from hard disk and SSD solutions; 2)Learn how multi-tenancy in a cloud datacenter relates to Persistent Memory security features; 3)Show how protection of data at rest can facilitate secure reuse of NVDIMMs.

The SNIA NVM Programming model enables applications to consume emerging persistent memory technologies through step-wise evolution to greater and greater value. Starting with an overview of the recently released revision of the NVM programming model specification version 1.2, this session summarizes the changes and the continuing work of the NVM programming TWG in areas of high availability and security. We take an application view of ongoing technical innovation in a persistent memory ecosystem.

Learning Objectives: 1) Learn about the exciting things happening in the SNIA NVM programming TWG; 2)Get an overview of several key topics of the SNIA NVM Programming Model; 3)Learn about the resources available to help developers plan and implement persistent memory aware software.

NVMe is one of the most interesting new developments to happen to storage in the past several years, and NVMe over Fabrics extends these capabilities over a Storage Area Network. Given that 80% of all existing Flash storage solutions deployed are interconnected with Fibre Channel (FC), many questions have arisen about what it is, how it works, and why someone might want to consider using Fibre Channel for NVMe-based solutions.

Learning Objectives: 1)How does Fibre Channel and NVMe work together?; 2)How is NVMe over Fabrics different from “traditional” Fibre Channel?; 3)What changes do I have to make to my Fibre Channel environment in order to get NVMe over Fabrics to work?; 4)What would an NVMe over Fibre Channel (FC-NVMe) architecture look like?; 5)What advantages would I get using Fibre Channel in an NVMe solution?

The industry needs a protocol that provides efficient storage operations for both data centers and hyperscale applications. Data centers are willing to pay more for more complex interfaces that provide reliable data response times, while Hyperscalers want the lowest cost and greatest flexibility to optimize their cost/performance ratios. Two paths are under consideration to provide deterministic read latency. One is I/O determinism in which the host controls the timing of reads vs. writes to storage elements specified by the controller. This approach requires complex controller firmware leading to highly capable but possibly more expensive controllers. Data centers find this to be easy to install and manage. The other path is Open Channel (Open Fabrics Alliance) where the controller provides information about the physical characteristics of the storage device (e.g., channel and die layout) and the host determines the physical location that data is stored on the medium. Hyperscalers believe this allows them to program their algorithms for writes and reads more effectively to prevent outlier read latencies, as they control all of the knobs on the device. There is also a perception that these SSDs should be less expensive. This configuration requires more initial effort by systems software and continued re-writing of code to keep up with changing SSD technologies that impact what criteria need to be controlled to provide optimal performance.

The implementation of SMB3 is a broad and important set of topics on the Samba roadmap. After a longer period of preparations, the first and the most generally useful of the advanced SMB3 features has recently arrived in Samba: Multi-Channel. This presentation will explain Samba's implementation of Multi-Channel, especially covering the challenges that had to be solved for integration with Samba's traditional clustering with CTDB, which is invisible to the SMB clients and hence quite different from the clustering built into SMB3. Afterwards an outlook will be given on other areas of active development like persistent file handles, RDMA, and scale-out SMB clustering, reporting on status and challenges.

Samba is used on a wide variety of local file systems, but it can also be used on cluster filesystems and now on NFS. This presentation will describe our experiences running Samba on NFS, with Samba as a gateway to pNFS storage, and the ways to best integrate the two protocols on Linux when run in this configuration.

The presentation will discuss how ACLs and SMB3 specific metadata can be handled, as well as some performance considerations.

A new class of ultra low latency remote storage is emerging - nonvolatile memory technology can be accessed remotely via high performance storage protocols such as SMB3, over high performance interconnects such as RDMA. A new ecosystem is emerging to "light up" this access end-to-end. This presentation will explore one path to achieve it, with performance data on current approaches, analysis of the overheads, and finally the expectation with simple extensions to well-established protocols.

Learning Objectives: 1) Understand the potential for low latency remote storage; 2) Survey the protocols and interfaces in use today; 3) See current performance data, and future performance expectations; 4) See a view of the future of the end-to-end storage revolution.

NVMe SSDs are becoming increasingly popular choice in scale out storage for latency sensitive workloads like databases, real time analytics, video streaming. NVMe SSDs provide significant performance throughput and lower latency compared to SATA, SAS SSDs. It is not unrealistic to expect these devices providing close to million random IOs per second. However scale out software stacks have significant amount of software overhead limiting the immense potential of NVMe SSDs. In this session, we present all flash scale out cluster performance, analysis on data path I/O overhead and programming techniques to systemically address software performance barriers.

Learning Objectives: 1) Scale out storage software data path flows; 2) Performance profiling with NVMe SSDs; 3) User mode v/s kernel mode NVMe SSD integration; 4) Optimization techniques.

Designed to meet the expectations of end users for simple, modern and secure management of scalable platform hardware, the DMTF’s Redfish is an open industry standard specification and schema that specifies a RESTful interface and utilizes JSON and OData to help customers integrate solutions within their existing tool chains.

This session provides an overview of the Redfish specification, including the base storage models and infrastructure that are used by the SNIA Swordfish extensions (see separate sessions for details). We will cover details of the Redfish approach, as well as information about the new PCIe and memory models added to support storage use cases.

Learning Objectives: 1) Introduction to Redfish concepts; 2) Application of REST APIs to standards management.