Starwind Software

Problem
Complexity and high implementation and maintenance costs slow down the adoption of virtualization in smaller SMBs and ROBOs. What works fine for Enterprises and larger SMBs just stops working with a smaller scale and multiple locations at some point, mostly because of required staff. The first enjoy a pleasant journey to construct the jigsaw puzzle of a “whole virtualization product” using state-of-art hardware and software components. They have trained personnel to design, assemble, test, fine-tune, maintain and support the resulting virtualization solution. Smaller SMBs cannot follow the same plan with their lean IT team or no full-time IT employees at all, while the sheer number of sites makes it impossible or very difficult for ROBOs as well. The first enjoy a pleasant journey to construct the jigsaw puzzle of a “whole virtualization product” using state-of-art hardware and software components. They have trained personnel to design, assemble, test, fine-tune, maintain and support the resulting virtualization solution. Smaller SMBs cannot follow the same plan with their lean IT team or no full-time IT employees at all, while the sheer number of sites makes it impossible or very difficult for ROBOs as well.
The situation is well-known, therefore there are different ideas on how to solve or work around, raising the following points:
  • Services to build and maintain SMB or ROBO virtualization project.
  • Pre-configured virtualization solutions shipped
  • VMs hosted off-site
However, most of them are not effective, because:
  • Services do not come free or cheap, so IT operational expenses actually increase instead of going down. It’s basically like taking money from one pocket and putting into the other – while virtualization cuts down some costs, others appear.
  • Pre-configured virtualization solutions do nothing with insane over-provisioning initially and poor scalability later “on the go”, resulting in bad ROI. Also ones coming from hypervisor vendors are just too expensive, and ones shipped by “budget” hyper-converged vendors typically use non-common components like rarely used hypervisors with very immature ecosystem around. The result is a serious vendor lock-in with support and maintenance difficulties.
  • Hosting business VMs off-site implies lots of the related IT operations being hosted as well. This is not always technically possible or reasonably priced for smaller business units in any case.
As a result, up to this date, there was no virtualization solution designed for lots of remote offices, smaller and especially very small SMBs.

Solution
For ROBOs, mid-sized and especially small SMBs who are looking to bring in quick deployment and operational simplicity to virtualization workloads and reduce related expenses, our solution is a Hyper-Converged Platform (H-CP). StarWind unifies: commodity servers and storage hardware, hypervisor and associated software into a single scalable layer. StarWind Hyper-Converged Platform has minimalistic initial hardware footprint of just two servers, called Hyper-Converged Platform Units (H-CPUs). Network switches are optional. Scaling unit is just one server at a time. Compute and storage resources can be leveraged individually matching customer workload and other requirements. H-CPUs are pre-configured, pre-tested and pre-tuned for every specific case, and shipped by StarWind’s hardware partners.
StarWind employs the “best of breed” components for H-CPUs: Servers from Dell, disks from WD, flash from Intel, etc. As remote offices and smaller SMBs typically use Windows, we chose Microsoft Hyper-V as the hypervisor. StarWind Virtual SAN is used for virtual storage and Veeam Backup and Replication utilized for VM backup. In order to manage the hypervisor, Microsoft SCVMM, or 5nine Manager are employed along with 5nine Antivirus for better security. We cut down operational costs even further by allowing the use of either a free Hyper-V Server or a full-blown Windows Server 2012 R2 with Hyper-V role enabled. The latter comes with a number of Windows VMs already licensed, thus reducing TCO. StarWind is the “one throat to choke” in terms of support. StarWind H-CP costs at least 40% less than all the components separately, thanks to cooperation between StarWind and their software and hardware partners.
StarWind also offers hardware-less H-CP for use on the customer’s existing set of hardware. It will take industry-standard hardware from the vendor of choice, software-only version of StarWind H-CP and reference design guide to deploy the Hyper-Converged Platform. In addition, it is possible to test the H-CP capabilities with a free software bundle.
Conclusion
StarWind Hyper-Converged Platform is a virtualization solution crafted right for SMBs and remote offices: not adopted or “corner cut” for them, but designed from scratch and purpose-built to meet their unique needs. It comes either as an all-inclusive turnkey product complete with H-CPUs or a software solution to deploy the H-CP on existing industry-standard hardware.
 
Hyper-Converged
As the name implies, StarWind HyperConverged Appliance adopts hyperconverged architecture, combining compute and storage resources into one scalable layer. It is exceptionally easy to manage and very cost-efficient, reducing all virtualization expenses, including energy consumption. For these reasons, Hyperconvergence is “the way to go” for SMB and ROBO with their limited budgets and understaffed IT teams.  

Converged (“Compute and Storage separated”)
StarWind HyperConverged Appliance is, obviously, hyperconverged. However, it can become a part of Converged architecture, serving storage outside the appliance. It uses industry-standard uplink protocols: SMB3, NFS and iSCSI, so it easily integrates into any virtualization infrastructure.  

VM-Centric Storage and StarWind Log-Structured File System (LSFS)
StarWind HyperConverged Appliance employs StarWind VSAN Log-structuring feature to handle virtualization workload. It battles the common virtualization-associated issues that cause performance degradation, like “I/O blender”. Thus, Log-structuring makes the minimalistic resources enough resources for even Enterprise routine, not to mention everyday SMB and ROBO tasks.  

Server-Side Cache
StarWind HCA uses multi-tiered RAM and flash caching for exceptional performance, reducing I/O latency to an absolute minimum. Utilizing StarWind VSAN, it achieves high performance with conventional SATA SSD, RAM and MLC flash. Caches are distributed between the nodes, achieving fast and reliable write back policy without employing uber-expensive proprietary hardware.  

Snapshots and Automated Storage Tiering
StarWind HCA employs StarWind VSAN inter-node tiering technology, which is essentially a performance-preserving and flash-saving mechanism. It offloads the rarely-used “cold” data from expensive primary storage to inexpensive secondary tier, stopping performance degradation and prolonging flash life.  

Deduplication and Compression
StarWind HyperConverged Appliance uses space reduction technologies for better performance and capacity utilization. Deduplication and compression reduce the actual amount of data to be transferred, written, read or processed, thus removing the unnecessary workload from HDD, flash and wire.  

Hardware Agnostic and Commodity Hardware
StarWind HyperConverged Appliance uses commercial off-the-shelf hardware, which is part of the reason for its cost-efficiency. There is no overpriced proprietary stuff, typically associated with turnkey solutions that make the customer pay extra for custom rails, etc. StarWind HCA is built on affordable commodity hardware that comes in a discount bundle, dramatically reducing the price of purchase and maintenance.  

Fault Tolerance and High Availability
StarWind HyperConverged Appliance achieves fault-tolerance and high availability with a minimalistic two-node setup. It utilizes StarWind VSAN to “mirror” storage and caches between the nodes. As a result, the appliance can withstand failures of individual disks, memory modules or even an entire node without compromising uptime.  

Scale-Up and Scale-Out
StarWind HCA scales according to the business needs, adopting both scale-out and scale-up architectures. Ready HCA nodes or individual flash and disks can be added to handle the growing workload without “overkill” overprovisioning and spending more money than is actually required.  

Asynchronous Replication
StarWind HyperConverged Appliance comes with an industry-leading backup solution – Veeam Backup and Replication. It will ensure the safety of mission-critical data and minimize the risk of data corruption even in case of disasters. There is also the option to use public or private cloud as the DR site, for example – Azure.  

Virtual Storage Appliance
There is a version of StarWind Virtual SAN, designed for fast deployment from a pre-configured VMware or Hyper-V compatible VM appliance. In case one can find a suitable application for this version in StarWind HCA, it is not prohibited. However, the Virtual Storage Appliance is meant to install on existing hardware set for testing purposes, rather than full-fledged production platform, mainly because it needs some tinkering by StarWind engineers to achieve maximum possible performance.
 
Virtual Tape Library (VTL)
StarWind HCA comes with a disabled VTL option for obvious reasons. However, in case the customer decides to use the appliance in his tape backup infrastructure, he can always enable it back. There are absolutely no restrictions on this matter.  


 - Prodata - Beyond Technology 

Company
Invalid Input

First Name(*)
Please type your full name.

Last Name
Invalid Input

Telephone
Invalid Input

Mobile:
Invalid Input

E-mail(*)
Invalid email address.

Type
Invalid Input

Description
Invalid Input

Product
Invalid Input

Invalid Input

First Name(*)
Please type your full name.

Last Name
Invalid Input

Company
Invalid Input

Telephone
Invalid Input

Mobile:
Invalid Input

E-mail(*)
Invalid email address.

Product
Invalid Input

Description
Invalid Input

Invalid Input