SNIA has a new specification in town – focused on key value storage.  SNIA on Storage sat down with Bill Martin, Co-Chair of the SNIA Technical Council and Co-Chair of the SNIA Object Drive Technical Work Group, to understand why SNIA took on this project and what are the results. SNIA On Storage (SOS):  Bill, thanks for taking the time to chat with us.   To get started, can you tell me what key value storage is and how it relates to the Technical Work charter that SNIA undertakes? Bill Martin (BM):  Key value storage is a new method of storing data when compared to the traditional block storage method.  You store a “Value” related to a “key (address)”, with the ability to then look up the value in the future using the “key” of the associated object. Within SNIA, the Object Drive Technical Work Group (TWG) has the charter to establish architectures and standards for disk-, solid state-, and tape drive- based functionalities that allow them to be higher level storage nodes in emerging scale-out solutions.  The TWG develops specifications that are vendor-agnostic and support existing and future functionality in drive form factors. Under the charter, the  Object Drive TWG has developed and published the Key Value Storage API Specification, a SNIA Technical Position that defines an application programming interface (API) for Key Value storage devices. With a storage device that enables key value storage, the API allows upper level applications to take advantage of the key value store method, independent of what infrastructure is being used.  The API outlines a set of functions that access key value storage the same way regardless of lower level protocol (NVMe, PCIe, or SCSI).  The provider of the lower level protocol provides a library interface that translates the key value API calls into whatever the lower level function calls need.  The API does the heavy lifting- freeing the application to not worry about what bus the storage is on. SOS:  Why did SNIA undertake this work? BM:  As storage is evolving from disk-based to non-volatile memory (NVM), there is a different paradigm for how data is addressed on media. Disk storage has logical and physical block address mapping. With NVM storage, the media is addressed differently; so for a key to be stored using the traditional mechanism, the host translates the key address to a logical block address, and then the storage device translates to the physical location on the media.  Therefore, if we can take a key as an address and hand it to the NVM storage device, we only need to do a single translation to store or retrieve the value on the media.  Eliminating the second translation reduces the time needed for mapping and storing the data on the physical device and a mapping table in the host – the “location of the data” now resides on the key value storage device, eliminating both physical storage and latency. SOS:  What work is being done to make Key Value Storage available for the operating systems? BM:  The first place the API will be used is in the Linux stack.  Samsung, a SNIA and Object Drive TWG member, has developed open source software for use in the Linux stack.  The drivers will be based on lower level protocol – a NVMe key value storage device. While Linux is the first effort by reason of the customer base focus on Linux-based operating systems, subsequently there will be drivers for Microsoft. SOS:  Is SNIA doing the lower level functional standards for key value storage as well? BM:  The SNIA API programming interface is independent of the lower level commands.  One lower level protocol interface is being developed by a technical work group within the NVM Express organization, and will be announced by them at a later date. Many of the same individuals working on the SNIA specification are also working on the NVM Express specification, taking advantage of a strategic alliance announced between the two organizations and DMTF in 2018 (Strategic Alliance Formed with SNIA, DMTF, and NVM Express) SOS:  Is there more work to do on key value storage? BM:  Version 1.0 of the Key Value Storage Specification is now out as a published SNIA Technical Position and available for anyone to use.  The Object Drive TWG has developed another published Technical Position on IP- based drive management.  Work now underway in the TWG is on a management interface for key value storage devices.  The TWG is looking at the need to manage key value storage devices to find out their capabilities and select the capabilities they are interested in to effectively utilize the API to move data.  Currently, the KV API Specification is a set of functions that allow storage and retrieval of data.   The management work will look at how to configure storage, incorporating work done by the SNIA Scalable Storage Management (SSM) Technical Work Group on the SNIA Swordfish^TM open storage management model as an extension of the DMTF Redfish system management model. SOS:  What do you see impacting storage and management in the future? BM:  I am very enthusiastic about the relationship between key value storage and computational storage.  Today what we can do is address storage based on a set of keys.  Because key value storage eliminates the extra key to block transaction, it will facilitate computation.  In the future, we will be able to request computation be done on the values associated with the keys and return only the meaningful data. SOS: Where can blog readers learn more about key value storage? BM:  SNIA has an exciting new Educational Library, which cross references over 700 pieces of SNIA content assets, including presentations, white papers, tutorials, webcasts, and technical specifications.  I’d first recommend reading the Key Value Storage Specification, and watching this video on Key Value Storage Standardization Progress. You can then search the library for other key value storage items, and related topics such as NVMe, object drives, and Linux.

The European Commission, Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs has issued COMMISSION REGULATION (EU) 2019/424 on 15 March 2019 laying down eco-design requirements for servers and data storage products pursuant to Directive 2009/125/EC of the European Parliament and of the Council and amending Commission Regulation (EU) No 617/2013. While the focus is energy-related requirements, embedded in this regulation is a requirement for servers and storage systems having 4-400 drives (systems with less or more are exempted) to have an ability to perform secure data deletions; this functionality is required from 1 March 2020. “Secure data deletion” means the effective erasure of all traces of existing data from a data storage device, overwriting the data completely in such a way that access to the original data, or parts of them, becomes infeasible for a given level of effort. SNIA’s Green Technical Working Group is preparing a guidance document associated with this regulation and are working through a variety of issues. The SNIA Security Technical Work Group (TWG) has already prepared a white paper on Data/Media Sanitization and how it related to ISO/IEC 27040 (Storage security). With this new regulation and the anticipated update of ISO/IEC 27040, the Security TWG will be developing further materials in this area. Click here for more information or to participate in SNIA’s Storage Security Technical Work Group, or contact SNIA’s Technical Council Managing Director at tcmd@snia.org.    

Issues related to security have great importance in IT today. SNIA is participating in the creation of international standards with leading security-focused industry organizations. Here’s an update on recent activities from the SNIA Security Technical Work Group (TWG): Transport Layer Security

• The SNIA Security TWG is keeping a keen eye on the TLS 1.3 landscape, which is starting to get interesting since the IETF approved RFC 8446 last August. TLS 1.3 is significantly different from previous versions and it is expected to have an impact on the mandatory elements for the SNIA TLS Specification for Storage and ISO/IEC 20648:2016, which are based on TLS 1.2. While TLS 1.2 is still valid and will be for some time, it is important to keep in mind that ISO standards like ISO/IEC 20648:2016 have a 5-year shelf life. SNIA plans to work on an update later this year so that a new specification is in place in 2021.

Storage Security ISO Standard

• The Security TWG completed its initial work on a potential refresh/update of the ISO/IEC 27040:2015 (Storage security) standard and submitted a recommendation to INCITS/CS1 (U.S. TAG to ISO/IEC JTC 1/SC 27) that included a proposal to subdivided it into at least 4 parts. The Security TWG also prepared 4 drafts and submitted along with the recommendation. SC 27/WG 4 accepted the U.S. recommendation at the April 2019 meeting in Israel and has initiated a Study Period on the revision of ISO/IEC 27040, which will leverage SNIA's draft and consider a fast-track approval process (joint NWIP and CD ballots on each document). The Study Period will present it recommendation at the SC 27/WG 4 meeting in Paris in October 2019.

Electronic Discovery

• With the inclusion of the Security TWG's text preservation, retention, and archive language from SNIA’s Data Protection technical white paper along with additional comments, ISO/IEC 27050-4 (Electronic discovery – Part 4: Technical readiness) has progressed to the Committee Draft (CD) stage. The Security TWG is now working with the 27050-4 editing team on specific text for the CD draft. This ISO/IEC 27050 standard targets both the legal and records management communities and Part 4 is very relevant to the storage industry.

Supply Chain Security

• The TWG is also seeing Supply Chain Security as a topic that will definitely come up in 2019. ISO/IEC JTC 1/SC 27 has initiated revision of three of the parts for the multi-part ISO/IEC 27036 (Supply chain security) standard and the SNIA TWG will be an active participant.

Security Techniques

• The TWG has been actively monitoring the ISO/IEC 27552 project (Security techniques — Extension to ISO/IEC 27001 and ISO/IEC 27002 for privacy information management — Requirements and guidelines) in SC 27, which has been approved for publication (expected availability in late summer 2019). Given the focus of the Europeans, we are reasonably certain that this standard will serve as the basis for organization certification for privacy (e.g., for GDPR). Efforts are underway to help the storage industry understand what it is and how it could affect us.

Other areas that the TWG is actively tracking

• AI, big data, IoT, smart cities, blockchain, etc. anticipating that some or all of them could have an impact especially in the computational storage arena
• Fabric security as it relates to NVMe

To learn more, please consider joining the SNIA Security Technical Work Group, visit https://www.snia.org/technology-focus/data-security or contact the SNIA Technical Council Managing Director at tcmd@snia.org.

For a long time, the architecture and best practices of storage networks have been relatively well-understood. Recently, however, advanced capabilities have been added to storage that could have broader impacts on networks than we think. The three main storage network transports - Fibre Channel, Ethernet, and InfiniBand – all have mechanisms to handle increased traffic, but they are not all affected or implemented the same way. For instance, utilizing a protocol such as NVMe over Fabrics will offer very different methodologies for handling congestion avoidance, burst handling, and queue management when looking at one networking in comparison to another. Unfortunately, many network administrators may not understand how different storage solutions place burdens upon their networks. As more storage traffic traverses the network, customers face the risk of congestion leading to higher-than-expected latencies and lower-than expected throughput. That's why the SNIA Networking Storage Forum (NSF) is hosting a live webcast on June 18, 2019, Introduction to Incast, Head of Line Blocking, and Congestion Management where our NSF experts will cover:

• Typical storage traffic patterns
• What is Incast, what is head of line blocking, what is congestion, what is a slow drain, and when do these become problems on a network?
• How Ethernet, Fibre Channel, InfiniBand handle these effects
• The proper role of buffers in handling storage network traffic
• Potential new ways to handle increasing storage traffic loads on the network

Register today to save your spot for June 18^th. As always, our experts will be available to answer your questions. We hope to see you there.

For a long time, the architecture and best practices of storage networks have been relatively well-understood. Recently, however, advanced capabilities have been added to storage that could have broader impacts on networks than we think. The three main storage network transports – Fibre Channel, Ethernet, and InfiniBand – all have mechanisms to handle increased traffic, but they are not all affected or implemented the same way. For instance, utilizing a protocol such as NVMe over Fabrics will offer very different methodologies for handling congestion avoidance, burst handling, and queue management when looking at one networking in comparison to another. Unfortunately, many network administrators may not understand how different storage solutions place burdens upon their networks. As more storage traffic traverses the network, customers face the risk of congestion leading to higher-than-expected latencies and lower-than expected throughput. That’s why the SNIA Networking Storage Forum (NSF) is hosting a live webcast on June 18, 2019, Introduction to Incast, Head of Line Blocking, and Congestion Management where our NSF experts will cover:

• Typical storage traffic patterns
• What is Incast, what is head of line blocking, what is congestion, what is a slow drain, and when do these become problems on a network?
• How Ethernet, Fibre Channel, InfiniBand handle these effects
• The proper role of buffers in handling storage network traffic
• Potential new ways to handle increasing storage traffic loads on the network

Register today to save your spot for June 18^th. As always, our experts will be available to answer your questions. We hope to see you there.

UCSD Computer Science and Engineering, the Non-Volatile Systems Laboratory, and the Storage Networking Industry Association (SNIA) are inviting submissions of proposals for presentation at the first annual Persistent Programming in Real Life (PIRL) conference.  PIRL brings together software development leaders interested in learning about programming methodologies for persistent memories and sharing their experiences with others. This is a meeting for developer project leads on the front lines of persistent programming; not sales, marketing, or non-technical management. PIRL is small, with attendance limited to under 100 people, including speakers.  It will discuss what real developers have done, and want to do, with persistent memory.  It will involve what worked, what didn’t, what was easy and hard, what was surprising, and what others can learn from the experience.  Presenters are encouraged, and even expected, to show and write code live in the presentation in a comfortable and dynamic peer environment. Possibilities for presentations include, but are not limited to: •  Experiences on a particular project •  Live code development showing new concepts •  Code challenges •  New tools for programming All attendees will be provided access to a development environment to respond to code challenges, or to show their own work in small forums.  This is intended to be a competition-free atmosphere for peers to network with each other to advance the use of persistent memory in the industry and academia.  By combining many of the industry leaders with the academic lights driving practical applications of new technology, peers at PIRL will encourage forward progress for adoption of persistent memory in the marketplace. Keynote speakers include key personnel from Dreamworks, VMWare, Oracle, Eideticom, and Intel. PIRL will be hosted by the Non-Volatile Systems Laboratory at the University of California, San Diego.  It will be held at Scripps Forum on July 22nd to 23rd, 2019, with optional events starting July 21st. Pre-registration will be $400. We’re excited to present this new conference, and we’re excited for you to participate.  Submit your presentation or code challenge idea today. Submissions are due by Monday, June 10th.  

UCSD Computer Science and Engineering, the Non-Volatile Systems Laboratory, and the Storage Networking Industry Association (SNIA) are inviting submissions of proposals for presentation at the first annual Persistent Programming in Real Life (PIRL) conference.  PIRL brings together software development leaders interested in learning about programming methodologies for persistent memories and sharing their experiences with others. This is a meeting for developer project leads on the front lines of persistent programming; not sales, marketing, or non-technical management. PIRL is small, with attendance limited to under 100 people, including speakers.  It will discuss what real developers have done, and want to do, with persistent memory.  It will involve what worked, what didn’t, what was easy and hard, what was surprising, and what others can learn from the experience.  Presenters are encouraged, and even expected, to show and write code live in the presentation in a comfortable and dynamic peer environment. Possibilities for presentations include, but are not limited to: •  Experiences on a particular project •  Live code development showing new concepts •  Code challenges •  New tools for programming All attendees will be provided access to a development environment to respond to code challenges, or to show their own work in small forums.  This is intended to be a competition-free atmosphere for peers to network with each other to advance the use of persistent memory in the industry and academia.  By combining many of the industry leaders with the academic lights driving practical applications of new technology, peers at PIRL will encourage forward progress for adoption of persistent memory in the marketplace. Keynote speakers include key personnel from Dreamworks, VMWare, Oracle, Eideticom, and Intel. PIRL will be hosted by the Non-Volatile Systems Laboratory at the University of California, San Diego.  It will be held at Scripps Forum on July 22nd to 23rd, 2019, with optional events starting July 21st. Pre-registration will be $400. We’re excited to present this new conference, and we’re excited for you to participate.  Submit your presentation or code challenge idea today. Submissions are due by Monday, June 10th.  

In the last few years, Ethernet equipment vendors have announced big increases in line speeds, shipping 25, 50, and 100 Gigabits-per -second (Gb/s) speeds and announcing 200/400 Gb/s. At the same time Fibre Channel vendors have launched 32GFC, 64GFC and 128GFC technology while InfiniBand has reached 200Gb/s (called HDR) speed. But who exactly is asking for these faster new networking speeds, and how will they use them? Are there servers, storage, and applications that can make good use of them? How are these new speeds achieved? Are new types of signaling, cables and transceivers required? How will changes in PCIe standards and bandwidth keep up? And do the faster speeds come with different distance limitations? These are among the questions our panel of experts will answer at the next live SNIA Networking Storage Forum (NSF) webcast on May 21, 2019, "New Landscape of Network Speeds." Join us to learn:

• How these new speeds are achieved
• Where they are likely to be deployed for storage
• What infrastructure changes are needed to support them

Register today to save your spot. And don't forget to bring your questions. Our experts will be available to answer them on the spot.

In the last few years, Ethernet equipment vendors have announced big increases in line speeds, shipping 25, 50, and 100 Gigabits-per -second (Gb/s) speeds and announcing 200/400 Gb/s. At the same time Fibre Channel vendors have launched 32GFC, 64GFC and 128GFC technology while InfiniBand has reached 200Gb/s (called HDR) speed. But who exactly is asking for these faster new networking speeds, and how will they use them? Are there servers, storage, and applications that can make good use of them? How are these new speeds achieved? Are new types of signaling, cables and transceivers required? How will changes in PCIe standards and bandwidth keep up? And do the faster speeds come with different distance limitations? These are among the questions our panel of experts will answer at the next live SNIA Networking Storage Forum (NSF) webcast on May 21, 2019, “New Landscape of Network Speeds.” Join us to learn:

• How these new speeds are achieved
• Where they are likely to be deployed for storage
• What infrastructure changes are needed to support them

Register today to save your spot. And don’t forget to bring your questions. Our experts will be available to answer them on the spot.

Thousands of IT decision makers, operators and the developers will gather April 29 – May 1 at the Open Infrastructure Summit in Denver, Colorado to collaborate across common use cases and solve real problems.

On Monday, April 29, from 2:50 p.m. – 3:30 p.m., members of the OpenSDS project and the Technical Working Group (TWG) which develops SNIA Swordfish™, are holding a Birds-of-a-Feather (BoF) session at the summit titled “Open Storage Management.”

To kick things off, Richelle Ahlvers, SNIA board member, chair of the Scalable Storage Management TWG, and storage management architect, Broadcom, will provide a brief overview of the SNIA Swordfish storage management specification. Swordfish is an extension to the DMTF Redfish^® specification that provides a unified approach for the management of storage equipment and services in converged, hyper-converged, hyperscale and cloud infrastructure environments. Swordfish is built using a RESTful interface over HTTPS in JSON format, and also provides support for OpenAPI.

Richelle will also discuss the lifecycle of creating consistent open standard interfaces, from definition to implementations, and how the open source ecosystem plays a role in open infrastructure management.

Xing Yang, principal architect at Huawei Technologies, and project and architecture lead in OpenSDS, will explain how the open source community addresses storage integration challenges in scale-out cloud native environments and connects siloed data solutions.

The session will be interactive and attendees will be encouraged to join in the conversation, get their questions answered and share their knowledge while making valuable new connections. Add the BoF to your conference schedule here.

While visiting the summit, stop by to see SNIA in booth #B13 in the Open Infrastructure Marketplace and pick up the latest Swordfish swag!