NVMe Hot-Plug Challenges And Industry Adoption

3y ago
38 Views
3 Downloads
587.35 KB
5 Pages
Last View : 4d ago
Last Download : 3m ago
Upload by : Baylee Stein
Transcription

Direct fromDevelopmentPowerEdge Product GroupNVMe Hot-Plug Challenges and Industry AdoptionTech Note by:Dell EMC PowerEdge 14th-generation (14G) servers support a wide varietyReliability, Availability, Serviceability, and Manageability (RASM) featuresdesigned to enhance server uptime and reduce total cost of ownership, asshown in Figure 1 below:Austin BolenSUMMARYDell EMC understands that hotplug operations for NVMe SSDswhile the server is running areessential to reducing andpreventing costly downtime.PowerEdge 14G servers supporta wide variety of hot-plugserviceability features, including:Surprise insertion, which enablesaddition of NVMe SSDs to theserver without taking the serveroffline.Surprise removal on OSes thatsupport it, which allows a user toquickly remove a faulty, damaged,or worn out NVMe SSD.Figure 1 Reliability, Availability, Serviceability, and ManageabilityOne notable RASM feature supported on PowerEdge servers is theserviceability of Hard Disk Drives (HDDs) and Solid-State Drives (SSDs) andmost recently NVM Express (NVMe) solid-state drives (SSDs). NVMe is anindustry standard storage protocol designed to optimize performance of solidstate drives. Serviceability features allow NVMe SSDs to be added, removed,or replaced without the server having to be opened or turned off. This allows foreasy replacement and/or re-provisioning.NVMe SSDs in the U.2 2.5” form-factor are typically located in the front ofPowerEdge servers which enables the easiest accessibility, however there aredesigns where these devices reside in the rear of the server. Refer to theInstallation and Service Manual for your PowerEdge server for more details onthe location and servicing of NVMe SSDs. 2019 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC and other trademarks are trademarks of Dell Inc. or its subsidiaries

Serviceability is further enhanced by allowing U.2 2.5” NVMe SSD mounted in the front or rear of the server to beserviced while the server is powered on and running using an industry feature referred to as hot-plug which maximizesavailability by minimizing costly server downtime. Hot-plug is broken down into two operations: Hot Insert: You insert an NVMe SSD into a running server.o Hot Removal: You remove an NVMe SSD from a running server.o Surprise Insertion: Prior to physically inserting the NVMe SSD, you do not notify the system that the NVMeSSD is about to be inserted.Surprise Removal: Prior to physically removing the NVMe SSD, you do not notify the system that the deviceis about to be removed.There are also orderly operations where operating system commands are used.oOrderly Insertion: Prior to physically inserting the NVMe SSD, you notify the system that the NVMe SSD isabout to be inserted.oOrderly Removal: Prior to physically removing the NVMe SSD, you notify the system that the NVMe SSD isabout to be removed.PowerEdge servers and the operating systems supported on them support surprise insertion. There is no need tonotify the system before hot-inserting an NVMe SSD.Note: For surprise removal of any storage device (SAS, SATA, USB, NVMe, etc.), the user must ensure the data isnot critical to the functioning of the system before removing the storage device. For example, a non-RAID boot storagedevice or swap file storage device could typically not be removed from a running system as doing so would likely crashthe operating system.Figure 2 Hot-Pluggable NVMe SSDs 2019 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC and other trademarks are trademarks of Dell Inc. or its subsidiaries

Direct fromDevelopmentPowerEdge Product GroupThe factors below also impact the ability to successfully hot-plug NVMe SSDs on PowerEdge servers: Form Factor – Hot-plug is only supported on U.2 2.5” form factor NVMe SSDs externally accessible in thefront or rear of the server. PCIe Adapter Card NVMe devices do not support hot-plug. Mechanicals – PowerEdge servers are designed with high insertion count connectors on our backplanedesigns as well as the NVMe SSDs we use. For hot insertions, the NVMe SSD needs to be fully inserted. Forhot removals, the NVMe SSD needs to be fully removed. Number of NVMe SSDs supported - All NVMe U.2 2.5” SSDs are hot pluggable, but only one at a timeshould be hot plugged. Operational Times – Hot-plug operations are only supported once the operating system has loaded. Hot plugoperations are not supported when the OS is shutting down or in pre-boot. Timing – Hot-plug operations should be performed in a timely manner. Removal and insertion should becompleted within 1 second.We’ve discussed above what hot-plug is and why it is important to users. We will now go into details on the interdependencies of the operating system and BIOS to support hot-plug operations with NVMe SSDs.For many storage device protocols, such as SAS, SATA, and USB, there is no need for orderly removal operationsprovided the data on the drives is not critical for continued operation of the system. For these protocols, surpriseremove will suffice. Many operating systems, NVMe device drivers, and applications may not support surprise removalof NVMe SSDs.Operating systems, drivers, and applications have many years of hardening to be able to reliably handle surpriseremoval of SAS, SATA, and USB storage devices. In all of these cases, there is a storage controller that acts as anintermediary between the storage device and the operating system, drivers, and applications. While the drivesthemselves are removed, the SAS, SATA, and USB storage controllers that the operating system, drivers, andapplications talk to remain in place and are never removed. These controllers are shown above the hot-plug barrier inFigure 3.Figure 3 Storage Controller Hot-Plug Comparison 2019 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC and other trademarks are trademarks of Dell Inc. or its subsidiaries

In NVMe, the storage controller was moved down on to the storage device below the hot-plug barrier as shown inFigure 3. An advantage of this approach is that it removes an added component layer when compared to the otherstorage solutions and helps NVMe to achieve such low latency accesses. However, this introduces a new model thatoperating systems, drivers, and applications had not dealt with before: the storage controller itself is removed when thestorage device is removed.Also note in Figure 3 that SAS, SATA, and USB have their own dedicated bus that have been architected for hot-plug.NVMe SSDs sit on the PCIe bus. The Conventional PCI bus architecture created in the 90s had no support for hotplug. Afterwards, a hot-plug model referred to as the Standard Hot-Plug Controller (SHPC) nt/download/8236) was added to Conventional PCI but requiredorderly removal and orderly insertion. When PCIe was introduced (the follow-on to the Conventional PCI/PCI-Xbusses) it adopted the SHPC orderly insert/remove model.There was some rudimentary support for surprise removal added to PCIe but it was not architected with thecomplexities of NVMe SSDs in mind. Only recently has there been strong market demand for hot-plug PCIe devices.As a result, operating system vendors and application developers have not invested much effort into supporting thisuse case. As new protocols emerge that benefit from PCIe surprise removal like NVMe and Thunderbolt, operatingsystems and applications developers are starting to add more robust support, but it will take time for all operatingsystems and applications to complete the PCIe surprise remove support. As of the writing of this paper, the followingoutlines the state of support for NVMe SSD surprise removal in various PowerEdge server supported operatingsystems: Microsoft supports surprise removal of NVMe SSDs starting with Windows Server 2016. VMWare operating systems do not yet support surprise removal of NVMe SSDs. Full support for surpriseremoval of NVMe SSDs is expected in a future release of ESXi. The Linux server distributors like Red Hat Enterprise Linux, SUSE Linux Enterprise Server, and Ubuntu Serverdo not yet support surprise removal of NVMe SSDs. There are many developers actively submitting patches tothe Linux kernel to harden the support for surprise removal of NVMe SSDs. Dell EMC continues to work withthe open source Linux kernel community and the Linux server distributors and expect these operating systemsto have full support for surprise removal of NVMe SSDs in future releases.Many aspects of the system need to be modified in order to support surprise removal of NVMe SSDs. Dell EMC hasmade the changes at the server level (BIOS/UEFI System Firmware, iDRAC, backplanes, cables, etc.) and to DellEMC applications/drivers (OpenManage Server Administrator, Dell Update Package, S140 Software RAID, etc.) tosupport surprise removal of NVMe SSDs. Dell EMC has also worked with the PCIe silicon vendors that provide PCIeroot ports and PCIe switches used in PowerEdge servers to ensure they support surprise removal of NVMe SSDs.Dell EMC qualified NVMe SSDs also support features needed for surprise removal such as power-loss protection(PLP) which ensures they can commit data in volatile memory buffers on the NVMe SSD to persistent memory on apower loss due to surprise removal or other conditions. When using NVMe SSDs not qualified by Dell EMC, the usershould check with the vendor of those NVMe SSDs to ensure they support surprise removal.For operating systems or applications that do not support surprise removal of NVMe SSDs, Dell EMC managementtools such as OpenManage Server Administrator and iDRAC provide the user with an option to do an orderly removalvia the “Prepare to Remove” task. Figure 4 on the following page shows the “Prepare to Remove” task for an NVMeSSD in OpenManage Server Administrator. For more details on the “Prepare to Remove” task refer to the User’sGuide for OpenManage Server Administrator and roducts/software int/software ent systems mgmt) and DellPowerEdge Express Flash NVMe PCIe SSD 2.5 inch Small Form Factor fsh-nvme-pcie-ssd users-guide en-us.pdf). These management tools will attempt to determine if the 2019 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC and other trademarks are trademarks of Dell Inc. or its subsidiaries

Direct fromDevelopmentPowerEdge Product GroupNVMe SSD is in use and warn the user if so. They cannot detect all cases where an NVMe SSD is in use and so theuser should verify the NVMe SSD is no longer in use prior to removing it. Some operating systems may preventorderly removal of NVMe SSDs that are still in use.Figure 4 Prepare to Remove NVMe SSDUsers will need to check with the vendor of any operating system or third-party application that accesses NVMe SSDsto determine if it supports surprise removal of NVMe SSDs. For operating systems or third-party applications that donot support surprise removal of NVMe SSDs, users should perform an orderly removal as described above.Dell EMC is also working with various industry standards bodies such as PCI-SIG (https://pcisig.com/) and the ACPISpecification Working Group (https://www.uefi.org/workinggroups), silicon providers, operating system vendors, andother OEMs to define new industry standard mechanisms to further improve support for NVMe hot-plug operations inthe future.ConclusionsDell EMC PowerEdge 14th-generation (14G) servers support a wide variety of hot-plug serviceability features for NVMExpress (NVMe) Solid-State Drive (SSDs) that address RASM and improve TCO. Surprise insertion is supported toallow adding NVMe SSDs to the server without taking the server offline. For operating systems that support it, surpriseremoval is supported to allow a user to quickly remove faulty, damaged, or worn out NVMe SSDs. Dell EMCunderstand that hot-plug operations for NVMe SSDs while the server is running reduces costly downtime and aredriving the industry to improve user experience. 2019 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC and other trademarks are trademarks of Dell Inc. or its subsidiaries

NVMe SSD is in use and warn the user if so. They cannot detect all cases where an NVMe SSD is in use and so the user should verify the NVMe SSD is no longer in use prior to removing it. Some operating systems may prevent orderly removal of NVMe SSDs that are still in use. Figure 4 Prepare to Remove NVMe SSD

Related Documents:

Austin Bolen, Dell EMC Myron Loewen, Intel Lee Prewitt, Microsoft Suds Jain, VMware David Minturn, Intel James Harris, Intel 4:55-6:00 8/7/18 NVMe-oF Transports: We will cover for NVMe over Fibre Channel, NVMe over RDMA, and NVMe over TCP. Brandon Hoff, Emulex Fazil Osman, Broadcom J Metz,

DPDK cryptodev Released In progress NVMe-oF Initiator BDEV NVMeoF BD NVMe-oF Target. 18. SPDK Virtual BDEV Perfect place to add storage algorithms SPDK NVMe NVMe-oF Target NVMe Driver BDEV NVMe BD SSD for Datacenter BDEV enables stackable SW BDEV provides abstraction for storage solutions to be inserted Storage Services can be:

Jun 30, 2020 · Q2. Q3. Q4. Q1. Q2. Q3. Q4. NVMe 1.2.1 May’16 Transport and protocol RDMA binding. NVMe-oF 1.0 May’16. NVMe-MI 1.0 Nov’15 Out-of-band management Device discovery Health & temp monitoring Firmware Update. NVMe 1.3 May’17 Sanitize Str

875319-b21 hpe 480gb sata ri m.2 2280 ds ssd 875587-b21 hpe 480gb nvme x4 ri sff scn ds ssd 875589-b21 hpe 960gb nvme x4 ri sff scn ds ssd 875591-b21 hpe 1.92tb nvme x4 ri sff scn ds ssd 875593-b21 hpe 400gb nvme x4 mu sff scn ds ssd 875595-b21 hpe 800gb nvme x4 mu sff scn ds ssd

the majority of M.2 SSD drives on the market are still AHCI based, and not NVME. An Example of an NVME based M.2 SSD drive is the Samsung SSD 950 Pro[4], shown in Figure 2. NVME drives typically use M.2 "type M" edge connectors, allowing them access to four PCIE lanes. The U.2 interface for NVME SSD drives allows traditional 2.5 inch physical form

**Comparison of models based on specifications provided on Crucial, Western Digital, and Kingston company websites as of September 9, 2021. MAINSTREAM SSD SEGMENT CS1030 PNY CS1030 CRUCIAL P1 WD GREEN SN350 KINGSTON NV1 Form Factor/Interface M.2 NVMe PCIe G3 x4 M.2 NVMe PCIe G3 x4 M.2 NVMe PCIe G3 x4 M.2 NVMe PCIe G3 x4

NVM Express (NVMe) is a protocol for the transport of data over different me-dia and for optimized storage in NAND flash. Peripheral Component Intercon-nect Express (PCIe) is currently the most used transport medium. Other media, like NVMe over Fabrics, are currently being standardized. NVMe is optimized for NAND flash chips.

THE GUIDE SPRING BREAK CAMPS 2O2O MARCH 16–27 AGES 5–13. 2 2020 Spring Break Camp Guide WELCOME Build Your COCA Camp Day 2 March 16–20 Camps 3–4 March 23–27 Camps 5–6 Camp Basics 7 Registration Form 8–9 Registration Guidelines/Policies 10 Summer’s coming early this year! Join us over Spring Break for unique and fun arts learning experiences. You’ll find favorites from .