Why Hyper-Converged Infrastructure (HCI)?
CIOs are looking for solutions that will simplify their data center operations, allowing seamless growth and reduced operational expenditures (OpEx), without sacrificing performance and security. HCIs offer vast improvements in scalability, performance, management and VM density over converged infrastructure solutions, which are often various pieces of disparate equipment in the same rack. Implemented in modules or nodes, hyper-converged systems are much simpler to expand as storage or compute power is required.
Gridstore HCA: Data Center in a Box
A single Gridstore all-flash HCA delivers the resources to run a data center – each appliance has four independent nodes with configurable cores, RAM, NICs and flash storage (select from configurations tailored to workloads).
Delivering superior performance for demanding graphics workloads, especially in a virtual desktop infrastructure (VDI), requires specialized GPU processing. In this graphic-rich age, high performance, low latency graphics processing increases end user satisfaction and productivity. Gridstore all-flash Graphics Nodes extend HCI and provide optimal price/performance for graphics-intensive processing.
Scaling that separates storage from compute, high performance built with affordable and reliable all-flash, erasure coding that eliminates replicas and a single infrastructure with single-pane-of-glass management lead to superior data center economics.
Take advantage of Windows Server 2012 R2 data services
Gridstore leverages its strategic partnership with Microsoft to support WS2012 R2 features, allowing Microsoft Windows Server users to increase their ROI by using the full set of features available from Windows Server.
Gridstore All-Flash HyperConverged Infrastructure
Gridstore all-flash HCI is a data center game changer. Virtual machines (VMs) cost less, run faster and consume less space, power and cooling compared to traditional and converged infrastructures. Gridstore all-flash HCI does for infrastructure what virtualization did for servers.
Gridstore’s unique architecture does not replicate the full infrastructure stack three times like other hyper-converged platforms. Instead, Gridstore uses erasure coding that offers the same fault tolerance and usable capacity but uses 50% less infrastructure to achieve this. This is a direct savings of 50% capital cost and 50% ongoing operating costs.
Not all HCI solutions are the same. Gridstore all-flash HCI outperforms local replica architectures with higher VM density and 50% lower TCO.
Purpose-Built for Windows Workloads and the Microsoft Cloud-Inspired Data Center
Gridstore HyperConverged Infrastructure includes:
- HyperConverged Appliance
Gridstore’s 2U HCA is the building block of the Gridstore HyperConverged Infrastructure solution. Each appliance has up to four HyperConverged Nodes that can be scaled to a maximum of 256
HyperConverged Nodes (compute/storage) or Storage Nodes (storage only). Its all-flash architecture delivers higher VM density, faster storage I/O performance and reduced power, cooling and management requirements.
- Windows Servers
Any Microsoft Windows Servers with or without Hyper-V to which are added the Gridstore vController, a software driver that interfaces with Gridstore HCAs and provides simple management and end-to-end control of I/O from server to storage, eliminating the “I/O blender effect” and allowing performance to scale linearly with capacity.
Gridstore All-Flash HyperConverged Appliance
The Gridstore all-flash HyperConverged Infrastructure solution transforms the data center, enabling higher virtual machine (VM) density and faster storage I/O performance, both underscored with significantly lower administration costs.
Gridstore HCAs deliver the following features and benefits:
|High performance erasure codes||
|Offload graphics processing from CPU to GPU||
|Always on; no downtime||
|Microsoft Windows Server 2012 R2 Data Services Support||
|End-to-end I/O control per VM
|Flexible configuration options
|Independent scaling of compute
|Native Windows kernel
** “Faster, More Cost-Effective VDI Deployments with All-Flash HyperConverged Infrastructure,” Gridstore, July 2015
*** Arun Taneja webinar, “Why Hyper-Convergence is Invading the Data Center,” October 16, 2014
Select From Configurations Tailored to Workloads
Meeting the performance and storage needs of Windows workloads is simplified with Gridstore HCAs, which include compute, storage, and storage networking all in a single system. Starting with as few as three nodes and scaling up to 256 with over 6PB of all-flash capacity, Windows workloads are supported with the power and storage that fits an IT budget, while allowing easy scaling up as needs change. Choose from these three options to achieve the most efficient, high performance, cost effective infrastructure with lowered upfront costs, ongoing lowered CapEx and OpEx, and VM optimization:
• HCA with dual processor E5-2690 v3, 32GB – 512GB DRAM (configurable) and 2x10GbE RJ45 ports
• HCA with dual processor E5-2640 v3, 32GB – 512GB DRAM (configurable) and 2x10GbE RJ45 ports
• HCA with dual processor E5-2620 v3, 32GB – 512GB DRAM (configurable) and 2x10GbE RJ45 ports
Selecting the right storage is also a key requirement for supporting Windows workloads. Gridstore offers three SSD options:
• 960GB SATA III 2.5” SSD for mixed workloads
• 1920GB SATA III 2.5” SSD for read-intensive workloads
• 3840GB SATA III 2.5” SSD for read-intensive workloads
In addition to scaling HCAs with SSDs, you can add external storage with Gridstore Storage Nodes.
Optimal Price/Performance HCI Extended With Gridstore Graphics Node
Gridstore HCI includes the ability to offload graphics processing from the CPU to the GPU in virtualized environments with its Gridstore all-flash Graphics Node (FGN), delivering superior performance for demanding graphics workloads, such as in medical imaging or CAD environments in a virtual desktop infrastructure (VDI), in addition to supporting big data computational and graphics requirements. In this graphic-rich age, true PC experiences are vital to virtual users. Productivity and end user satisfaction are improved with high performance, low latency graphics in a Gridstore all-flash HCI environment.
Supporting up to two FGNs in each Gridstore HyperConverged Appliance (HCA), and allowing compute to scale separately from storage, Gridstore delivers the flexibility required for today’s changing business needs. Each Gridstore HCA can support all-flash HyperConverged Nodes (FCN), all-flash Storage-only Nodes (FSN), all-flash FGNs, or any combination of these. There are three GPUs to choose from in the FGN which can be paired with any of the supported CPUs in the Gridstore product line.
HCI Built for Graphics Processing – Choose From Three Options
Gridstore has a unique architecture for HCI that drives significant competitive advantages and real customer benefits over all other HCIs. The Gridstore architecture uniquely combines three innovations to drive breakthrough density, economics, and price-performance.
Breakthrough Density with Flexible Scaling
Hyper-converged systems combine compute, storage, and storage networking into a single tier. Gridstore’s HyperConverged Appliance delivers high compute power combined with all-flash storage and up to four of these independent nodes in a single 2U system.
Superior Flexible Scaling
The Gridstore architecture addresses one of the biggest issues with HCI, which is combining compute and storage together as the scaling unit. In effect, this forces customers to scale compute at the same rate as storage, which typically grows at 5-6X the rate of compute. This turns the advantages of being very simple to deploy into the very high cost of compute sprawl.
The Gridstore architecture is designed to allow you to scale compute and storage independently as well as leverage your existing infrastructure to eliminate the creation of another infrastructure silo, eliminating the cost of compute sprawl.
- Scale Storage: Our bare metal architecture allows us to scale storage independently by adding separate appliances that do not require a full compute stack hypervisor or licensing on top of it. We reduce the amount of CPU and memory required just to run the resource controller without any other overhead. This allows you to grow storage pools to fit whatever requirement you have.
- Scale Compute: Our bare metal architecture also means we can place a virtual controller (vController) onto any host, whether physical, virtual or your existing hosts. The vController has the same access to an all-flash storage infrastructure as a HyperConverged Node.
Another key aspect of Gridstore’s unique architecture is how we scale. Instead of replicating the full infrastructure stack three times like other hyper-converged platforms, Gridstore uses erasure coding that offers the same fault tolerance and usable capacity but uses 50% less infrastructure to achieve this. This is a direct savings of 50% capital cost and 50% ongoing operating costs.
Gridstore is the only HCI with a parallel all-flash architecture. This chart shows the difference between an all-flash HCI with a hybrid storage-based HCI. Gridstore delivers the highest performance of any hyper-converged system on the market with an all-flash infrastructure that, due to our architecture, we can deliver at less cost than hybrid storage-based alternatives.
For detailed information on the Gridstore HCI architecture, read this whitepaper.
Leveraging Microsoft Windows Server 2012 R2 Data Services
Gridstore leverages its strategic partnership with Microsoft to support WS2012 R2 features, such as the following. This allows Microsoft Windows Server users to increase their ROI by using Microsoft’s full set of features.
• Cluster Support – Gridstore provides SCSI block devices that support Cluster Shared Volumes (CSV) to a single Windows Sever, Hyper-V hosts, or Microsoft Scale-out File Clusters.
• Seamless Management per VM – Storage is easily managed from the GridControl MMC snap-in and storage services like snapshot, replication are managed through standard Microsoft tools such as Hyper-V Manager on a per VM basis.
• Storage QoS – Gridstore TrueQoS™ is Hyper-V aware so the I/O requests from each virtual machine are prioritized, based on the priority of the application to an organization, end to end, right down through the physical storage.
• Microsoft Hyper-V – Gridstore uses its unique software defined storage architecture to optimize the I/O performance of each virtual machine and then takes advantage of its H-Class Storage Nodes using flash as a cache to deliver the IOPS required for demanding Hyper-V environments or databases.
• SMB3 – Gridstore complements SMB3 with block storage, and takes advantage of the impressive advances made by Microsoft in its file storage stack and SMB3 protocol. The result helps assure an easier-to-administer, highly scalable and 100% compatible storage system.
GRIDSTORE GAINS MICROSOFT GOLD STATUS FOR DATA CENTER COMPETENCY