Windows Storage Spaces Direct is known for continuous availability and high scalability for a much lower cost than many of the other options on the market. It creates software defined storage by using servers with locally attached drives. The architecture can be designed to be either converged or hyper-converged, making procurement and deployment a lot easier. Its amazing levels of efficiency and performance are boosted by its main features like storage tiers and caching, in addition to its use of the latest hardware like NVMe drives.
As mentioned, Storage Spaces Direct can be designed for two deployment options, converged and hyper-converged. With the converged option, compute and storage are in different clusters. Whereas with the hyper-converged option, the storage and compute are together in one cluster. Both have their advantages, but converged infrastructure is generally more flexible than hyper-converged infrastructure as it allows for easier scalability and more efficiency.
There are certain hardware requirements that arise when you choose to use Storage Space Direct. For networking hardware, the software is designed to use SMB3 instead of ethernet. For storage hardware it supports anywhere between 2 and 16 servers with locally attached NVMe, SAS or SATA drives. For each individual server, there needs to be 2 SSDs and 4 extra drives as a minimum requirement. If you use SAS or SATA, then the devices need to be behind an HBA (host-bus adapter) and SAS expander.
Since 2012 there are have been a large number of improvements, innovations and newly added features that have made Storage Space Direct extremely effective.
Storage Replica is a feature that allow for volume replication between servers. It is very effective for disaster recovery and gives you the relief of knowing that you have an extra layer of protection between your data and a catastrophe. Using the reliable and proven technology of SMB3, this feature is extremely effective. Storage Replica also has all of the advanced characteristics of SMB3 available to it.
The Software Storage Bus is a new feature that has essentially replaced Shared SAS cabling. It creates a software-defined storage fabric through which every server can see the local drives of the other servers. The Software Storage Bus is a feature that will dynamically join the fastest possible present drives to the drives that are lagging behind. By doing this it provides caching that speeds up IO and increases throughput. The Storage Pool feature creates a storage pool that is made up of a collection of drives. Any drive that is eligible is added to the pool automatically.
There is also a feature that provides a good amount of fault tolerance to the virtual "disks". It's done by either mirroring, erasure coding, or a combination of the two. The Resilient file system is a system built from scratch for virtualisation, increasing both the acceleration of key file operations and built-in error detection/correction system.
Volume sizes up to 64TB and single file sizes up to 1TB are now supported by Data Deduplication, which wasn't the case prior to 2016. You are also now able to easily back up data to the cloud using a Microsoft Azure-based cloud service. The VHDX files that are hosted on ReFS volumes get significant performance boosts due to the Resilient File System. The work folders feature means any changes in files are automatically synced directly to the work folders server. Storage quality of service feature (QoS) can be used centrally observe and monitor the end to end performance of your storage. Powershell comes with a huge advancement in new features, including class development support, more security capabilities, an improved usability, and more. Both virtual and physical machines will gain from the latest time synchronisation feature. They will have a much higher time accuracy due to the effects of the improvements in both Win32 Time and Hyper-V Time Synchronisation Services. Services will be compliant with the potential regulations that will require a minimum of a 1ms accuracy in relation to UTC. Finally, the JEA feature (Just Enough Administration) is a security technology that allows delegated administration for essentially anything that you can choose to manage through Windows PowerShell.
Boasting constant availability, Storage Spaces Direct has built-in resiliency features to deal with failures. Its high level of fault tolerance is due to its resiliency in handling component failures and maintaining its availability. Should any hardware fail, you just remove it and put in a replacement. There is no complex management process needed, the software makes the necessary adjustments itself.
Storage Spaces Direct is incredibly scalable. You can scale down to 2 servers, and scale as high as 16 servers and more than 400 drives. That makes up to 1 petabyte per cluster. Scaling is simple, all you do is add more drives or more servers. The software automatically onboards the latest drives and they will immediately become in use. The level of performance will always scale in a reliable and consistent manner.
Once deployed, there are a number of different maintenance tasks you will likely perform over the long term. These generally include such tasks monitoring for and fixing issues, replacing any parts that do fail, and updating your drivers.
There are four common tools that you can use to manage Storage Spaces Direct. Windows Admin Centre and Windows PowerShell are two of the most common. System Centre Virtual Machine Manager & Operations Manager is another, while the final common management option is Failover Cluster Manager & Server Manager.
Windows Admin Centre is a powerful tool, with a decent level of ease of use. It needs to be installed, and once done so it has helpful displays that allow you to monitor the status of individual servers and clusters.
Windows PowerShell is a great option if you are looking to automate and want full control over every single aspect of storage. The only real negative is that its better for management, rather than visual monitoring of system health. In that aspect, Windows Admin Centre is a more effective option for management.
System Centre VMM & Operations Manager is one of the easiest management tools to use and also provides a nice graphical view of the status of multiple systems. It's a common tool and widely used for management of Storage Space Direct. The only down side could be occasions where tasks that require PowerShell pop up, meaning this tool won't be enough to cover all situations. There are also times where extra infrastructure may become needed.
Failover Cluster Manager & Server Manager is similar in a way. It is also easy to use and has good graphical displays. However, like the previous option it also may require additional tools like PowerShell for certain tasks.
Broadberry is the premier provider of Storage Spaces Direct storage servers. With decades of experience in this field, we have built an impressive track record of reliability, consistency and high standards of excellence. Over the years we have supplied servers to many of the very biggest organisations on the planet including the BBC, SKY, NASA, TOYOTA. TESCO, SONY and many more. Having won awards for our service, we pride ourselves on delivering the very best customer experience possible.
Our storage servers are not only fully compatible with Storage Spaces Direct, but they are compatible with any storage server software. We are the only company that provide this level of compatibility across the board. Ordering a storage server from us will save you big money in the long term as if you ever want to change software or upgrade to something new, you won't be forced to purchase a whole new server. With us one box is all you need to run any software you could ever want to run. We are proud to be the sole providers of storage servers with such widescale compatibility.
In addition to this, our storage servers are also configurable with third party hardware. Unlike the majority of other storage server suppliers in this market, we do not believe in vendor locking. You will have complete freedom to use whatever hardware you wish to use. Many companies trap you into only buying hardware from them and their partners. We will never do this.
Our prices are as fair as they can be. We provide servers with reliability, performance and quality on par with very best in this industry. However, unlike many of the big-name brands we do not mark-up our products to exorbitant prices. We aim to provide the highest quality servers for the most affordable price, and even at lower prices our systems regularly outperform more expensive options.
With us, there is one price. The price we agree on for the initial purchase and any long-term maintenance required. Unlike other companies in this field we will never hit you with hidden charges, meaning your TCO (total cost of ownership) will be much lower than you would get with the alternative brands on the market.
The Broadberry CyberStore WSS range includes Storage Spaces Direct in Windows Server 2019.A new feature of SSD is nested resiliency, which allows a two-server cluster to resist multiple hardware failures simultaneously without loss of storage availability, enabling users, apps and VMs to continue running uninterrupted.
There are two new resiliency options offered by Storage Spaces Direct in Windows Server 2019 implemented in software, without the requirement of any specialised RAID hardware:
Nested two-way mirror In each server, local resiliency is provided by two-way mirroring. Greater resiliency is then provided by two-way mirroring between the two servers. With two copies in each server, it is essentially a four-way mirror. Nested two-way mirroring delivers uncompromised performance, writes goes to every copy and read come from any copy.
Nested mirror-accelerated parity Essentially combine nested two-way mirroring and nested parity. In each server, local resiliency for the majority is provided by a single bitwise parity arithmetic, with the exception of new recent writes which utilise two-way mirroring. Then, further resiliency for all data is provided by two-way mirroring between the servers.
Capacity efficiency is the ratio of usable space to volume footprint. It's used to describe the capacity overhead that can be attributed to resiliency and will depend on which resiliency option you choose.
For example, storing data without any resiliency is 100% capacity efficient (as 1 TB physical storage capacity). In comparison, two-way mirroring is 50% efficient (1 TB of data will take up 2 TB of physical storage capacity).
Nested two-way mirror writes four copies of everything. If you want to store 1 TB of data, you need 4 TB of available physical storage capacity. While a very simple option, the capacity efficiency of nested two-way mirror is only 25%. Meaning it has the lowest capacity efficiency of any resiliency option in Storage Spaces Direct.
Nested mirror-accelerated parity can achieve higher capacity efficiency, up to 35%-40%. That depends on two key factors. Firstly, the number of capacity drives present in each server. Secondly, the mix of mirror and parity you specify for the volume. This table below provides a lookup for commonly used configurations.
|Capacity drives per server||10% mirror||20% mirror||30% mirror|
You may notice that the capacity efficiency of classic two-way mirroring (which is about 50%) and nested mirror-accelerated parity (up to 40%) aren’t too different. Depending on what your requirements are, it may be worth taking the lower capacity efficiency to enjoy a serious increase in storage availability. As you can select resilience per-volume, you can mix nested resiliency volumes and two-way mirror volumes within the same cluster.
Simplicity. From industry-standard servers running Windows Server 2019 to your first Storage Spaces Direct cluster in under 15 minutes.
Unmatched Performance, featuring hypervisor-embedded architecture and built-in read/write cache.
Fault Tolerance. The built-in resiliency deals with drive, server, or component failures with continuous availability. Software heals itself, with zero convoluted management steps.
Resource Efficiency. Erasure coding provides up to 2.4x higher storage efficiency, done whilst minimising CPU consumption to send resources back to where they are most needed - the VMs
Manageability. With Storage QoS Controls you can keep overly busy VMs in check with minimum and maximum per-VM IOPS limits.
Scalability. Go up to 16 servers and over 400 drives, for up to 1 petabyte (1,000 terabytes) of storage per cluster.
Software Storage Bus The Software Storage Bus is a new feature in Storage Spaces Direct. Spanning the cluster, it establishes a software-defined storage fabric where every server can see the local drives of each other server. It has essentially replaced costly and restrictive Fibre Channel or Shared SAS cabling.
Storage Bus Layer Cache The Software Storage Bus dynamically binds the fastest drives present to slow drives, in order to provide server-side read/write caching that accelerates IO and boosts throughput.
Networking Hardware Storage Spaces Direct utilises SMB3, including SMB Direct and SMB Multichannel, over Ethernet in order to communicate between servers.
Storage Pool The storage pool is the collective drives that form the basis of Storage Spaces. It's created automatically, all eligible drives are automatically discovered and added to it.
Cluster Shared Volumes All ReFS volumes are unified into a single namespace by the CSV file system. This is accessible through any server, so to each server every volume looks and behaves as if its mounted locally.
Scale-Out File Server Only converged deployed need this final layer. It delivers remote file access using the SMB3 access protocol to client, like another cluster running Hyper-V, over the network. This essentially turns Storage Spaces Direct into network-attached storage (NAS).
Failover Clustering Built-in clustering feature of Windows Server is used to connect the servers.
Storage Hardware From 2 to 16 servers with local-attached SATA, SAS or NVMe drives. Each server requires at least 2 solid-state drives and at least 4 additional. The SATA and SAS drives should be behind a host-bus adapter (HBA) and SAS expander.
Storage Spaces Storage Spaces delivers fault tolerance to virtual disks by utilising erasure coding, mirroring or both. It can essentially be thought of as distributed, software-defined RAID utilising the drives in the pool. In Storage Spaces Direct, these virtual disks usually possess resiliency to two simultaneous drive or server failures (for example 3-way mirroring, each data copy in a different server) though chassis and rack fault tolerance is also available.
After you've deployed your storage, some of the management tasks you'll perform include monitoring for failed physical disks or other issues, replacing disks that fail, and updating firmware and drivers. See the following table for common management options.
|Windows Admin Center||-Powerful and easy to use
-Graphical displays for monitoring individual clusters and servers
|-Must be installed (it is pretty easy to install though...)|
|Windows PowerShell||- Complete control over all aspects of storage
- Can automate by writing scripts
- Powerful example scripts available
|- Better for management than visually monitoring overall storage health
- Scripts require development and testing
|System Center Virtual Machine Manager (VMM) & Operations Manager||- Easy to use
- Graphical displays for monitoring multiple systems
- Alerting for common problems
- Common tool for managing VMs and storage
|- Some tasks require Windows PowerShell
- Requires System Center licenses
- Might require additional infrastructure if you don't already have System Center or are deploying at a scale that's greater than your existing deployment can handle
|Failover Cluster Manager & Server Manager||- Easy to use
- Graphical displays for monitoring
|- Some tasks require Windows PowerShell
- Slow automatic refreshes in Server Manager when working with storage
- Performance and usability decreases with large numbers of objects
Low Powered servers from £470
Perfect Appliance Servers the CyberServe range of Intel Atom based rack servers are designed for light processing tasks. Designed and built for the appliance server market where an extremely low power consumption of less than 10W is required and 100% server up-time is of up most importance. Configurable with up to 64GB DDR4 RAM and 16 processing cores.
Single Processor servers from £707
Superb Business Class Servers, the CyberServe Xeon E range is based on Intel Xeon processors. An ideal fit for companies looking for an affordable and efficient system, these single socket servers can be configured with up to 64GB DDR4 RAM and 6 processing cores.
Single Processor servers from £939
Excellent Business Class Servers, the Broadberry CyberServe SP1 range of Single Xeon rackmount servers is based on the Intel Xeon Scalable Processor family. Providing huge benefits in performance, efficiency and security, these single socket servers can be configured with up to 512GB DDR4 RAM and 22 processing cores.
Dual Processor servers from £1,136
Perfect Enterprise-Class Servers the CyberServe SP2 Dual Xeon rack mount range is based on Intel Xeon Scalable Processor family. Delivering significant benefits in performance, power efficiency, virtualisation, and security. Servers are configurable with up to 3TB DDR4 RAM and 56 processing cores
Quad Processor servers from £681
Excellent Enterprise-Class Servers the CyberServe Quad Xeon rack mount server range is based on Intel Xeon Scalable Processor family. Capable of reaching scorching fast speeds, the solutions in this quad socket range can thrive in even the most computationally intense environments and deliver real speed to your data centre.
Single Processor servers from £1,209
Perfect Data Centre Servers, powered by single AMD EPYC 7002 series processors, this range reduces licencing costs by up to 50%. Providing up to 64 cores, 4TB of memory, and 128 lanes of PCIe 4.0 connectivity delivering double the I/O performance over PCIe 3.0, single-socket EPYC servers satisfy most of your workload needs.
Dual Processor servers from £3,010
Perfect Data Centre Servers, powered by dual AMD EPYC 7002 series processors, this range boasts the I/O, memory, and bandwidth to satisfy your needs. Providing up to 64 cores, 4TB of memory, and 128 lanes of PCIe 4.0 connectivity delivering double the I/O performance over PCIe 3.0, dual-socket EPYC servers provide market-leading performance.
Single Processor servers from £727
Excellent Business Class Servers, the CyberServe XE3® range is based on Intel® Xeon E3-1200 v6 processors. A perfect fit value-conscious companies searching for an effective and affordable system, CyberServe XE3® range is built for use as an entry level / mid-level server. Servers in this range are configurable with up to 64GB DDR4 RAM and 4 processing cores, with a single processor.
Dual Processor servers from £960
Perfect Enterprise-Class Servers the CyberServe XE5 2600 rack mount range are based on dual Intel Xeon E5 2600 v4 processors. Delivering significant benefits in performance, power efficiency, virtualisation and security. Servers are configurable with up to 1.5TB DDR4 RAM and 44 processing cores.
Single Processor servers from £1,205
Perfect Data Centre Servers, powered by single AMD EPYC 7001 series processors, this range boats performance at a fantastic price. Providing up to 32 cores, 8 memory channels and 128 PCIe 3.0 lanes per CPU unlocking capabilities and performance previously available only in 2-socket architectures.
Dual Processor servers from £3,105
Perfect Data Centre Servers, powered by dual AMD EPYC 7001 series processors, this range boats balanced architecture for optimum performance. Providing up to 64 cores, 4TB of memory, and 128 lanes of PCIe connectivity providing the flexibility, performance, and security features to efficiently support demanding workloads
Storage Spaces Direct uses industry-standard servers with local-attached drives to create highly available, highly scalable software-defined storage at a fraction of the cost of traditional SAN or NAS arrays.
Its converged or hyper-converged architecture radically simplifies procurement and deployment, while features like caching, storage tiers, and erasure coding, together with the latest hardware innovation like RDMA networking and NVMe drives, deliver unrivaled efficiency and performance.
2U, 24 Drive Bays, SAS / SATA Rackmount Storage Server
2U, (8 x SAS & 4 x NVME Bays) High-Performance Storage Server
The Broadberry CyberStore range of enterprise-grade storage servers are designed and built for a wide-range of storage applications.
Completely customisable from processing to storage type and capacity, to operating system - the CyberStore range is your go-to storage platform.
The CyberStore range of storage servers are used by the world's largest organisations from the BBC, to the CERN project, governement institutions, universities and much more.
Year-after-year, the Broadberry CyberStore range of unified storage appliances has beaten the likes of Dell and HPE to be crowned "Best Storage Appliance" by PC Pro magazine.
When compared to the likes of DELL, HPE, EMC etc, the Broadberry CyberStore range of enterprise-grade unified storage appliances is up to 60% cheaper, although using the same quality components such as Seagate drives and Kingston RAM.
The Bradberry CyberStore range has been built from industry-leading commodity hardware, and carefully designed to ensuure full compatibility with leading storage operating systems and database engines.
Broadberry enterprise storage appliances are on average 86% lower cost than EMC and 55% lower cost than DELL although built using the same leading-brand server components such as Intel processors and Samsung memory. HPE hard drives are rebadged Toshiba enterprise-class drives, the exact same models as found in Broadberry storage appliances.
When you buy a Broadberry storage appliance all drive bays are supplied with caddies - so if in 6 months time you decide to expand the storage you can source it freely from the marketplace or ourselves if you wish. The only way to add hard drives to HPE servers is to buy the drives along with caddies from HPE for 200% the price of Broadberry's price.
Broadberry storage appliances come with enterprise-grade features like IPMI included, whereas HPE charge large fees for perpetual licences for the same features.
Our website allows you to configure your server online with accurate prices, as opposed to NetApp who require several face-to-face meetings with them before they'll give you a price.
Over recent years there's been a massive move from the world's largest organisations such as Google to purchase their servers from the likes of Broadberry rather than DELL and HPE. This move is fuelled by the rise in commodity hardware.
Before leaving our UK workshop, all Broadberry server and storage solutions undergo a rigorous 48 hour testing procedure. This, along with the high-quality industry leading components we use ensure all of our server and storage solutions meet the strictest quality guidelines demanded from us.
Our main objective is to offer great value, high-quality server and storage solutions, we understand that every company has different requirements and as such are able to offer un-equaled flexibility in designing custom server and storage solutions to meet our clients' needs.
We have established ourselves as one of the biggest storage providers in the UK, and since 1989 supplied our server and storage solutions to the world's biggest brands. Our customers include: