MegaCast = MegaInformation

ActualTech Media recently hosted their first Hyperconvergence MegaCast to help people compare six different approaches to hyperconverged solutions and I presented the Gridstore solution. It was an informative three hours. You can listen to their whole replay – get a mega cup of coffee first! Next time attend live for a chance to win their cool giveaways.

For your convenience, clickHEREfor my megacast presentation.

DBAs stunned at Gridstore HCI performance

SQL Server deployments rely on fast performance and database administrators (DBAs) are tough to please when it comes to performance. Relying on traditional infrastructure is the norm, but with the changes with SQL Server licensing, the cost can be jaw dropping. Luckily, there’s a new norm in the makings. Hyper-converged infrastructure (HCI) allows virtualization of SQL Servers to enable faster deployment of new databases, more responsive client support, and the flexibility virtualization offers, but can it still meet DBA performance demands?

We have participated in bakeoffs where DBAs are convinced that the only way to ensure the highest level of performance is to stick with traditional infrastructure. Designing workload tests representative of a specific environment, and specifically designing a test that is intended to break the system, whether HCI or traditional, we have seen it all. In a recent bakeoff, the DBAs, who manage over 2,000 databases, ran their tests in parallel to put the most stress on the systems possible and see what the results were. Out of four systems tested, two broke and two passed the stress test.

Gridstore completed the workload and stress tests, as did a traditional all-flash array with SQL running on physical hosts—with Gridstore delivering about the same performance as the all-flash array. In the end, Gridstore proved to the DBAs the performance could be delivered while both consolidating the environment and reducing the layers of infrastructure through hyper-convergence to both reduce cost and greatly simplify the environment.

Read the full details on the customer’s environment, changes proposed and the results: SQL Server Consolidation with Gridstore HCI

Interview on HyperConverged Infrastructure with Henry Baltazar from The 451 Group

We recently had the opportunity to sit down with Henry Baltazar, Research Director for Storage at 451 Research, and ask a few questions around the Hyperconverged Infrastructure marketplace. Henry’s been writing about HCI for some time and we wanted to get his perspective on how the market is developing.

What has caused the rise of HCI?

Most organizations have been working to improve the speed and efficiency with which they deliver IT services, by overcoming perceived bottlenecks in how IT resources are deployed. Hyperconverged infrastructure (HCI) are being positioned as a means for achieving this goal, and vendors in this space claim their offerings can bridge gaps across resource silos to simplify the allocation process as well as day-to-day maintenance and troubleshooting.

Hyperconvergence has been receiving a tremendous amount of attention as the next step in the integration of IT resource delivery options. The technology takes the ideas of integrating compute, storage and networking that started with converged systems designs and added additional levels of abstraction and automation. The result is a concept that leverages virtualization and in-chassis storage to create systems from similar server-based building blocks. They promise simplified operation and the ability to expand by adding additional modules.

What problems do existing legacy infrastructure carry?

Storage administrators are being overwhelmed by data and workload growth and have struggled to keep pace with client demands.  The most disruptive aspect of HCI, is the rapid and simplified implementations they can facilitate.  We’ve had snapshots, replication, and other storage features for decades, what is different about HCI is that it gives IT generalists and virtualization administrators access and control of these technologies without requiring expensive proprietary arrays and SAN expertise.

What problems are customers trying to solve?

Provisioning time, for VMs, storage and networking resources, is a huge issue for enterprise and midsized organizations.  Client expectations continue to rise, and business stakeholders have little patience for the conventional IT resource allocation process.  Beyond provisioning time, conventional IT infrastructures with storage, server and virtualization silos, create management complexity and make problem resolution a nightmare for many organizations.

Are there particular organizations or verticals that are finding HCI most interesting, and why?

Right now, HCI’s sweet spot is in the midrange market.  Customers in this segment are not encumbered by deep relationships with enterprise storage vendors, and have IT staffs that favor generalists over specialists.  The relatively easy deployment of HCI and the ability to start out small and scale up are key attributes that make HCI popular with midsize customers.  The adoption of HCI has been broad to date, and there are a number of reference customers in verticals such as healthcare, government, manufacturing, and financial services.

Going forward we definitely expect to see an increased adoption for HCI in the enterprise market.  The Remote Office/Branch Office (ROBO) market in particular is key area where HCI can be beneficial for enterprise customers, since those offices and sites often do not have skilled IT staff and require IT infrastructure that can be easily deployed and remotely managed and maintained for the organization’s IT staff at the main offices.


What workloads benefit the most from HCI and why?

HCI offerings today are well suited for a broad market, and not really limited to specific use cases.  This is a big part of the appeal of HCI, and reduces the risk for customers since they know they will be able to use HCI in many different ways as their environments and workloads continue to evolve.  Early deployments of HCI focused on specific pain points such as VDI, where high performance and VM level management and monitoring granularity where challenging issues for traditional storage.  After proving themselves in the VDI use case, HCI vendors have been able to sell their offerings into database and other application use cases.  HCI has also been deployed in video surveillance and other media use cases, and we expect to see more product developments in this space to greatly expand the capacity of HCI offerings in the future.  It much easier to move an app to data, than it is to drag 100’s of TB or even Petabytes to an application server.

Historically HCI has not been a big player for high end Tier 1 use cases, which demand extreme performance and reliability.  This is changing however, as more HCI vendors bring all-flash systems to market, and with the addition of key data protection feature such as synchronous replication.  

What features and attributes should customers look for when shopping for an HCI offering?

There are a couple of areas I think customers should focus on:

-        Operational simplicity – How easy is it to provision resources and solve issues when they arise?

-        Agility and scalability – How fast can I deploy and optimize the HCI offering? Can I upgrade without disrupting the cluster?

-        Performance –How well does the HCI platform utilize flash storage and other key resources?  Can performance scale up with cluster size?

-        Resilience – Does the HCI offer features to ensure greater availability and protection from hardware failures?


Henry Baltazar is a Research Director for the Storage channel at 451 Research. Henry returned to 451 Research after spending nearly three years at Forrester Research as a Senior Analyst serving Infrastructure & Operations Professionals and advising Forrester clients of datacenter infrastructure technologies, including storage virtualization, cloud storage, solid-state storage and primary storage arrays. Henry has evaluated and tested storage hardware and software offerings for more than 15 years as an industry analyst and as a journalist.

Prior to his work with 451 Research and Forrester, Henry spent nearly nine years working as a technical analyst for eWeek Labs, where he covered storage, server hardware and network operating systems. At eWeek Labs, he initiated the testing coverage of various technologies, including data replication, clustering, virtual tape libraries, storage virtualization, SAN management, NAS, iSCSI and email archiving. In addition, Henry was a member of eWeek’s editorial board and provided content for the magazine’s enterprise storage blog. Henry has been widely quoted in the press, including such media outlets as Silicon Valley Business Journal, Computerworld and Henry holds a BA in Environmental Sciences from the University of California, Berkeley.


Webinar with 5nine Software: Q&A and Survey Results

Our latest webinar, presenting the Gridstore HyperConverged Infrastructure and 5nine Cloud Security solution, was another hit with our attendees (93% were highly satisfied/satisfied). We were asked many questions during the webinar and while we answered what we could, there were still a number that we didn’t get to. We’ve put together the answers to those questions as well as the results of our polls.

If you weren’t able to attend, we highly recommend you watch the webinar replay to find out how your environment can benefit from this solution (and you also have a chance to win an Amazon gift card). CLICK HERE FOR REPLAY

Below are the unanswered questions, along with the webinar survey results.

Q: What is the minimum size I can start at?
A: The minimum configuration, offering a full cloud-in-a-box solution, is the Gridstore HyperConverged Appliance (HCA) and three nodes.

Q: What ballpark price are we talking for a basic unit?
A: Gridstore can be deployed for around $35K, which includes an HCA and three nodes. If you want to demo a unit or set up a proof of concept, we will be happy to arrange that.

Q: Is RDGateway and/or NVGRE included?
A: 5nine Cloud Security supports all Windows Server networking configurations, including RD Gateways and NVGRE.

Q: Do I need separate licensing for the AV or IDS vendors?
A: No, all licenses are already included in your 5nine license, which for IDS includes Cisco Snort for Business, and for AV includes the choice of Bitdefender, Kaspersky Labs or ThreatTrack.

Q: Is there any special licensing options for Service Providers?
A: Yes, 5nine Cloud Security has a Service Provider License Agreement (SPLA) with monthly or annual billing at the “per VM” level. This allows hosting and service providers to pay for only what their end customers are using, allowing for fluctuations in usage.

Q: Do I need to use System Center or are there other options?
A: Both Gridstore and 5nine Cloud Security can be managed via Microsoft System Center but it is not a requirement.

Q: Will legacy devices slow down flash array?
A: Gridstore allows you to integrate existing Windows Servers in the same grid as the HyperConverged Appliance and take advantage of the HCI.

Q: The security, with little footprint, if it’s breached, what’s the protocol on your behalf (for example casinos)?
A: 5nine Cloud Security protects its own security system through redundancy, user tracking, change logging, and role-based access control. It also centrally monitors the health of each 5nine security component in the environment, and sends email or triggers other actions when an alert is triggered.

Q: Any plan for certification under SAP/HANA?
A: 5nine Cloud Security logs it events in the Syslog format, an industry standard. It supports any database that can use Syslog, including SAP HANA.

Q: Why align yourself with only MS and Hyper-V?
A: As the industry’s fastest growing hypervisor, Microsoft Hyper-V offers some unique capabilities and advantages over other solutions. Notably it supports the leading hybrid cloud solution through Microsoft Azure, Microsoft workloads (SQL Server, SharePoint, etc.) run faster on Hyper-V, and it is more affordable, especially since most enterprises are already paying for Windows Server in the guest OS and the host OS is free with Hyper-V Server 2012 R2.

Q: How does 5nine integrate into Hyper-V and make the management of Gridstore and Hyper-V overall easier and safer?
A: 5nine Cloud Security integrates in the virtual switch on each Hyper-V host. This enables agentless protection, as no security component is required to run inside each VM, and the network packets are intercepted and inspected before they arrive at the VM. This improves security, performance and management, enabling customers to run more VMs on each host.

Q: On an average how much overhead is needed for 5nine?
A: There is minimal processing overhead on each host, but this is overall more efficient doing it here than for each individual VM. Additionally 5nine’s patent-pending disk scanning technology runs up to 70 times faster than other solutions, minimizing the impact of scanning storms, which significantly increases the solution’s overall performance.

Q: Does 5nine require a separate Hyper-V cluster to ensure up time or can it exist in fault tolerance without a separate cluster?
A: 5nine Cloud Security can run on a single standalone VM or server. However it is recommended to put it inside a clustered VM for high-availability (this should be a standard cluster, it does not have to be dedicated). Multiple 5nine Cloud Security management servers can also be configured for redundancy, either locally or at branch offices.

Q: How much RAM and CPU do I need?
A: It depends on the number of hosts and VMs that you are trying to protect. Most customers use 2 – 4 GB of RAM, and larger customers will use up to 8 GB. All of this can run on a single CPU.

Q: Does it work with containers or docker for Windows?
A: These technologies are coming out with the future Windows Server 2016 release. 5nine’s intention is to support Windows Containers, along with Nano Server and nested virtualization, however our official support statement will not be available until closer to release.

Q: How do you build your threat library and how often are the virtual switch libraries updated?
A: 5nine leverages the IDS library from Cisco Snort for Business, and offers customers the choice of the antivirus library from either Bitdefender, Kaspersky Labs, or ThreatTrack. The user can control the frequency of updates on each host, which can be as often as hourly. Additionally a proxy server can be configured to pass updates locally, to minimize network traffic through public connections.

Q: Do you use signature or heuristics detection? Definitions can be outdated and let in malware.
A: Both. 5nine has its own heuristics that will map your datacenter’s regular traffic patterns and alert you about anomalies. 5nine leverages signatures for IDS from Cisco Snort for Business, and offers customers the choice of the AV signatures from either Bitdefender, Kaspersky Labs, or ThreatTrack.

Q: Any chance it also does deep packet inspection?
A: Yes, 5nine Cloud Security does DPI on network packets that are unencrypted using the HTTP protocol. We will be adding more protocols in the near future.

Q: What’s the timeline between a newly discovered zero day intrusion to a new threat library update to all customers?
A: Since 5nine Cloud Security leverages 3rd party IDS from Cisco Snort for Business, and AV from Bitdefender, Kaspersky Labs, or ThreatTrack, it is dependent on when they have a signature for the new threat. 5nine Cloud Security can update its libraries on demand or automatically as frequently as once per hour.

Q: Symon, I believe in the v7 release you have the capability of having a VM go into a specific group upon deployment. Can you expand on that or show how that is done?
A: Correct – you can create a default security group, and all newly created VMs will go to that group. These VMs will then be automatically and immediately protected using the policy for that group.

Q: Does the agentless configuration provide active protection within the VM operating system?
A: Active protection is an advanced alerting feature that will scan the inbound network packet for viruses. This requires an agent to be installed within the VM and is a great approach for static (infrastructure) VMs, rather than dynamic VMs used by end-users.

Webinar Attendee Survey Results
During the webinar, we presented a few polls to help us tailor the presentation, where appropriate. The following provides the results of the polls – check them out and see how you compare with our attendees.

Poll 1: How large is your Hyper-V production environment?


Poll 2: Which workloads are you considering for HyperConverged Infrastructure? (Select all that apply)


Poll 3: What role best describes your primary job function?


Webinar with Unidesk: VDI Survey Results and Q&A

Posted by Kelly Murphy, Gridstore and Tom Rose, Unidesk

We had a great time the other day presenting on a very interactive webinar with Dave Kawula, Microsoft MVP. The topic was Breakthrough Economics for VDI – how Gridstore HyperConverged Infrastructure coupled with Unidesk application layering technology and Microsoft Remote Desktop Services (RDS) enable virtual desktops to be implemented for far less than the cost of physical PCs – $550 per user “all-in.” We were asked many, many questions during the webinar and while we answered a lot, there were still a number that we didn’t get to. We’ve put together the answers to those questions as well as the results of our polls.


If you weren’t able to attend, we highly recommend you watch the webinar replay. In addition to us presenting and answering questions, Dave provided a live demo of the Gridstore/Unidesk VDI solution. Watch the full webinar to find out how your environment can benefit from this solution (and you also have a chance to win an Amazon gift card). CLICK HERE FOR REPLAY

Below are the unanswered questions (again, we answered many during the webinar which you will hear in the replay), along with the results of our VDI polls.

Q: We are looking to replace 50 desktops with a virtual replacement. We use Hyper-V for our server virtualization. Is it more expensive if I have less than 1500 desktops?

A: Both of us have configurations for smaller implementations, starting as small as under 200 desktops. There are volume breaks for Unidesk and Gridstore, but the price even for a few hundred desktops will be less than buying new PCs (<$800).


Q: Will mobile users have the ability to access their desktop through a single connection?

A: Yes. In all VDI systems, mobile users use the connection broker component to authenticate themselves and access their desktops through a single connection. Most connection brokers now support all client devices (e.g. PCs, thin clients, smartphones, pads). When using Microsoft RDS, Unidesk and Gridstore for VDI, the Microsoft Remote Desktop clients and the RD Connection Broker that are part of the RDS service in Windows Server 2012 will give your mobile users access to their desktops.


Q: What is the feedback/acceptance of VDI from the “true” end-users (non IT people, ie. those in sales, finance, HR, etc departments)?

A: VDI user acceptance ranges, of course, based on the level of service you provide. If you give them a locked down desktop that performs poorly and doesn’t save their customizations, it won’t be good. If you give them an all-flash Gridstore desktop that is faster than their old PC, and you save all personalization with Unidesk layers, and you can deliver apps in minutes instead of weeks, and you can repair their desktop in 2 minutes just by rebooting with Unidesk personalization layer rollback, they will love it.


Q: Is Unidesk another service/application that you include in your current environment? Or does it replace say a current Citrix environment?

A: Unidesk is application and image management software that you run in your current environment. If you are running Citrix XenApp or Citrix XenDesktop, you would install the Unidesk virtual appliance on your existing virtual infrastructure and use it to deliver layered applications and operating systems to XenApp session hosts or XenDesktop desktops. If you are running Microsoft VDI, you would install Unidesk in your Hyper-V environment and use Unidesk alongside Microsoft RDS.


Q: What is the bandwidth per machine needed for remote offices or do you need a caching server at each location?

A: VDI in general works better over high bandwidth networks, but WAN optimizations in the most common protocols – RDP/RemoteFX, HDX, PCoIP – now offer excellent performance even over lower bandwidth networks. You can also add WAN optimization solutions such as Citrix NetScaler, or leverage the distributed management architecture of Unidesk and Microsoft RDS to install a local Gridstore appliance in each remote office. The local users will then connect to the local office instead of the main datacenter, but you’ll still be able to manage all desktops, images, and apps once centrally.


Q: In our environment, if malware is detected we are required to wipe the HDD and re-install the image. How does VDI handle redirected folders or persistent layers?

A: Unidesk can eliminate malware and viruses simply by rolling back the desktop to an earlier snapshot of the Personalization layer. That’s one of the benefits of layering – OS, App, and Personalization layers are versioned. You can also “undo” bad patches simply by assigning an earlier OS or App layer version to one or more desktops and restarting them.


Q: With Unidesk’s single layer (disk image) architecture for OS & similarly for applications, storage IOPS demand on those layers (disk images) are expected to become a bottleneck. Do you have a recommendation (thumb rule) on design of the underlying storage platform…?

A: It’s actually the opposite. Unidesk is proven to bring out the best performance in hybrid storage arrays because the virtual disk layers shared by many VMs become “hot,” and stay resident in flash, which has more than enough IOPS to satisfy read/write requests. On Gridstore, Unidesk layering is even faster because of the all-flash architecture.


Q: How does the ‘Personal’ layer of Unidesk interact with User Profile Disks from MS?

A: Unidesk is compatible with Microsoft User Environment Virtualization (UE-V) and standard roaming profiles. Some customers create personal, persistent desktops with Unidesk and choose not to use UE-V or profiles, allowing the Personalization layer to capture all customizations – settings, data, user-installed apps, plug-ins, etc. Other customers create non-persistent, pooled desktops with Unidesk that are shared by many users – in this case, the Personalization layer is destroyed after each desktop use, so UE-V or profiles are used to restore user settings.


Q: Why not use Parent Disks for OS/Apps and Differencing Disks for the top layer?

A: Many of Unidesk’s 1300 customers tried that in their first attempts at VDI. Differencing disks get blown away every time you patch, which blows away all user personalization. That quickly became a non-starter for most VDI projects, and is why everyone is now looking to use layering.


Q: Is there any plan to also support the Linux and UNIX OS for this layering technology?

A: Not today. Unidesk is for Windows app mobility, at least in the near term.


Q: Is this pricing annual or monthly?

A: We showed the one-time cost for RDS CALs, Unidesk licenses and the Gridstore appliance. There will be the usual annual maintenance (16-20%) on top of that.


Q: What is the yearly maintenance cost after making the initial purchase of $550 / user?

A: Standard 16-20% based on quantity of licenses purchased.


Q: Is he saying that all software licenses are included with your pricing? To clarify, all software needed to run VDI – Hyper-V, Microsoft VDI Client Bundle etc.

A: We did not include the cost of Windows Server/Hyper-V because many Microsoft customers already have that as part of their EA. We also didn’t include the Microsoft VDA (virtual desktop access) subscription of $100 per user per year. That’s only needed if you are going to access the virtual desktop through a non-Windows device. If you are reusing Windows PCs, that cost is not needed. We were trying to keep it an apples-to-apples comparison of VDI vs. a fixed PC.


Q: Do you have a product for the 10-50 user range? We service many small businesses.

A: Great question. This solution can scale down quite nicely.


Q: What about the hardware platform supermicro – will Gridstore enter in discussion to be Dell-aware or Lenovo-aware?

A: Great question – we have the ability to separate software from hardware and are considering future platforms. You’ll receive an announcement from us when a new platform has been certified.


Q: Does Gridstore have plans to offer “compute upgrades” as processors advance, without replacing the entire unit?

A: We find most customers will just purchase a new node. Even though a chassis holds up to four nodes, you can grow by a single node at a time. You do not need to purchase an entire block of 4 nodes each time.


Q: What is the lifespan of this solution?

A: VDI typically has a longer lifespan than PCs – 5-7 years.


Q: Does the cost per user scale up?

A: Yes, it actually gets less expensive the more you buy.


Q: Do Unidesk & Gridstore come as an appliance only or will customers have the option of choosing underlying hardware from their partner of choice and use the Unidesk/Gridstore layer on top of this? ?

A: The solutions are independent. Unidesk can run on any platform, but you can only get this kind of performance / ROI on the Gridstore hardware platform.


Q: Any thoughts on the processing overhead for running the VDI? Is there an advertised number from MS for Hyper-V?

A: This is a great question. Normally the bottleneck in our testing comes down to CPU Cycles. We will see the CPU maxing out on the Login VSI tests long before the storage / network.


Q: CPU dropped after 100%. Was the test conducted in “High Power” mode?

A: The first and last of the graph are warm-up and cool-down. It’s all run in “High Power” mode.


Q: What happens when the Internet goes down?

A: If the solution is set up to use private networks or a local LAN in the same facility, the Internet going down will have no impact, as all traffic is local. If you are using the Internet for connectivity and it goes down, the RDP protocol will try to reconnect. If the session times out, the user would simply re-logon once the Internet was back online and they would be placed back onto their desktop, picking up exactly where they left off.


Q: Would Gridstore be able to manage/handle other type of containers? (Docker, LXD, etc.)

A: Gridstore doesn’t care what kind of data is stored on it, and is fully compatible with all container technologies. However, when it comes to running VDI workloads on Gridstore, only application layering with Unidesk and the provisioning and collection management extensions that Unidesk has added to RDS are going to be applicable. Also note that container technologies such as Docker capture the operating system together with the application and are isolated from all other containers, whereas Unidesk layers capture only the application’s registry keys and files and are fully interoperable with other layers (i.e. layered apps look and feel as if they’ve been locally installed).


Q: VDI cost doesn’t include the physical device at the user (Monitor/Keyboard/NIC/CPU). How much is a low-end physical device for a VDI environment?

A: There are some very good options out there for Thin Clients from IGEL Technology –> They have a soft client that can even turn an older PC into a thin client which really drops the cost. There are also micro-PCs in the $100 range that ship with Windows, enabling you to avoid the VDA subscription.


Q: Why didn’t VDI cost didn’t include physical endpoint?

A: Good catch – you were listening carefully! :-) For this model, we were assuming reusing existing PCs for virtual destkop access using IGEL, Dell, ThinPC or other “dumbing down” technologies others chatted about above. If you add in new thin clients and then the VDA license cost, that will bump the cost up a little, but it will still be less than new PCs.


Q: Are the solid-state drives deduped?

A: The Vstores can be deduplicated with Windows DeDupe and that scenario is supported by Microsoft + Hyper-V


Q: What about security? Working in healthcare there is always the HIPAA and HITECH regulations that need to be addressed.

A: By its nature, VDI enables better compliance with HIPAA and HITECH because patient data is no longer on the edge, but in the data center. VDI also provides a “follow me” desktop that lets clinicians pick up where they left off, promoting consistent use of EMR and other clinical systems. In addition, Gridstore and Unidesk are partnering with 5Nine to provide advanced security. Several joint healthcare customers are already using the Unidesk + Gridstore + 5nine combination – all better together.


Q: How do Gridstore and Unidesk work in a multi-tenant environment?

A: Both Gridstore and Unidesk are ideal for multi-tenant environments. Gridstore’s linear scaling makes it well-suited for small or large service providers who are offering hosted desktops for customers. Unidesk enables the Windows operating system and applications to be patched once to update all desktops. RDS collections, managed by Unidesk, can be used to separate tenants from each other and assign custom app layers on top of the same base OS layer. Highly granular roles/rights are being added in the next major release of Unidesk to allow administrators to “see” and manage only the applications and desktops used by their tenants.

The following provides the results of the polls taken during the webinar. Check them out and see how you compare with our attendees.


Poll 1: Have you got VDI deployed today?


Poll 2: What challenges concern you about VDI?



Poll 3: How many VDI users do you need to support?



We hope you enjoyed the webinar and this recap. Look for more developments from Gridstore and Unidesk as we execute on our recently announced partnership.

Unidesk, Microsoft and Gridstore: VDI doesn’t get any better than this!

Today we announced a partnership with Unidesk that will extend the reach of VDI worldwide. With amazingly simple deployment via Gridstore’s all-flash HyperConverged Appliance (HCA) and Unidesk’s application layering technology that extends Microsoft Remote Desktop Services (RDS), this partnership delivers a VDI solution that gives end users powerful performance for their VDI workloads, and IT managers a way to stretch their budget without sacrificing business needs.

Other “all in” VDI solutions are offered by solution providers, such as VMware and Citrix, but at a higher per desktop cost. Here is a cost comparison chart that shows the value of the Unidesk, Microsoft and Gridstore solution compared to VMware and Citrix, as well as traditional physical PCs – this partnership reduces the cost per desktop to $600!

  Physical PCs Gridstore



VMware Citrix
Number of desktops 500 licenses PC desktop management software 500 Microsoft RDS CALs & 500 Unidesk licenses 500 licenses Horizon Enterprise and 1 license vCenter Server 500 licenses Citrix XenDesktop & 500 RDS CALs
Cost (with reused PCs) $415,000 $300,000 $418,750 $437,500
Cost per year $830 per user $600 per user $837.50 per user $875 per user


What is not shown in this chart is the extreme performance the solution offers. With a solid foundation of affordable all-flash storage with inline erasure encoding in the Gridstore HCA, every VDI deployment immediately benefits from high I/O operations per second (IOPS) with sub-millisecond latency. Coupled with high performance software from Unidesk and Microsoft, VDI end users will get to experience the performance and responsiveness of an All-Flash Desktop.

There’s more to this solution that high IOPS and low cost and latency. It’s important to also point out the extreme scalability of this solution. Catering to data centers small and large, this VDI solution, with the Unidesk image management and application delivery capabilities, fits any IT budget, supporting from 100 desktops to tens of thousands in a true pay-as-you-scale model – simply by deploying additional VDI building blocks that deliver predictable performance and cost per desktop.

Tom Rose, Vice President Product Marketing, Unidesk, along with Dave Kawula, Microsoft MVP, will be joining me for a live webinar where we demonstrate how together Gridstore HyperConverged Infrastructure and Unidesk Virtual Desktop Management deliver breakthrough economics for VDI. I hope you can join us too.

Join our live webinar on Thursday, November 12th

(9am PST / Noon EST / 5pm GMT / 6pm CET)

Interesting Poll Results Around VMware to Hyper-V Migration

We recently presented a webinar topic that was of great interest to many people – how and why to convert from VMware to Hyper-V. During the webinar, we polled attendees about their migration plans. Surprisingly, some people attending had no intention of migrating! Not surprising was that the bulk of the people who intended to migrate from VMware to Hyper-V stated that lack of resources and experience were the main obstacles they faced. Fortunately, there are a myriad of migration tools available that reduce the amount of resources required, plus the business case for migrating shows that actually paying for consultants to do the conversion can be worth the upfront additional cost (due to all the VMware licensing costs).

With a response rate of over 50% from the 200+ attendees, these poll results are representative of the industry. The full set of polling questions and their results are:




To hear the full webinar, click here.
To access a great document loaded with migration tools, prepared by co-presenter John Kelbley, Cloud Specialist, Microsoft World Wide EPG, click here.

Design Goal #5: Scale to Fit Your Exact Requirements

Since the arrival of HyperConverged Infrastructure, the biggest customer complaint has been inefficient scaling. First-generation products force customers to scale servers at the rate of storage growth. To make matters worse, three-way replication force customers to grow servers at 3X the storage growth rate. For some workloads that are uniform (such as VDI), this is not a problem and many installations have started with these workloads. For many other general purpose workloads, however, storage grows considerably faster than the compute workloads do.

Unlike other HyperConverged Infrastructures, Gridstore is not iSCSI, NFS, or SMB serving I/O.

Rather, Gridstore is a peer-to-peer Grid architecture designed specifically for this purpose, addressing the following critical scenarios:

  1. Independent Scaling: Independently scale storage only using purpose-built storage nodes that can be deployed incrementally as storage is required. This supports mixing different classes of available storage (flash, hybrid, and capacity) to allow different types of capacity pools to grow to meet different workload requirements or budgets.
  1. Compute Scaling: Scale compute only using any X86 server that runs Windows. A Gridstore vController can be deployed in less than a minute to any Windows server. That server, in turn, can become a node in the Grid, with full access to high performance parallel all-flash performance to accelerate any workload regardless of the server it runs on. Hosts can be physical servers or hypervisors. If they are physical servers, features like end-to-end storage QoS can be applied to deliver precise amounts of IOPS to each of these servers. Hence, the Gridstore architecture allows scaling of both compute and storage separately. Compute can be scaled using any or existing X86 servers while storage can be scaled to fit precise requirements using storage-only nodes.

Gridstore to Partner with Scalability Experts for SQL Server Practice

A formal announcement between our two companies will follow, but we have just agreed to a partnership with Scalability Experts, These folks have built a successful and rapidly growing business focused on being data management and business intelligence experts. They offer scalability assessments, performance tuning, migrations and upgrades, consolidations, virtualization, BI and cloud computing services. They help their customers increase the performance of database operations, lower IT costs and improve business decision-making capabilities. Besides being a Microsoft National Systems Integrator (NSI), Gold Certified Partner and a four time Partner Award Winner, just this week they were recognized again here at the Microsoft Worldwide Partner Conference as Data Platform Winner.

Scalability Experts and Gridstore plan to integrate services and hardware into a packaged solution to bring advanced analytics to the Healthcare industry and to also help customers further consolidate and virtualize their datacenter environment.  Together, Gridstore and Scalability Experts will be able to bring solutions to customers that are affordable, scalable and will provide the highest performance.

If you are at the Microsoft Worldwide Partner Conference this week, come see us at booth #1106 and we’ll tell you more.


Design Goal #4: All-Flash HyperConverged Infrastructure

Efficiently and cost effectively deliver the highest possible performance

Because 66% of storage resources are wasted on three-way replicas, all-flash hyperconverged appliances have been cost prohibitive. Gridstore’s next-generation architecture is the first and so far only design that eliminates this problem and makes all-flash HyperConverged Infrastructure practical.

Gridstore uses flash in an end-to-end architecture to both accelerate performance and to eliminate as much as 80% of storage traffic, returning CPU cycles to hosts. In a future release Gridstore All-Flash HyperConverged Nodes will use flash as a local cache and as a high-performance storage resource. By putting a large flash cache (1TB) close to the workload, the entire working set of approximately 30 server VMs or 200 desktop VMs can be cached locally. Our vController sits in front of this cache, becoming part of a controlled data path with VM granularity. This critical feature ensures the cache resources are utilized according to per-VM policies and not consumed just by the most active VMs. With Gridstore, approximately 80% of I/O (reads) is served from the local low-latency flash cache. This eliminates 80% of storage traffic that would otherwise go across the network and allows storage nodes to focus on serving the remaining 20% (writes). By comparison, an all-flash array must handle 100% of the I/O (reads/writes) in the same controller through the same network connections. Gridstore’s end-to-end distributed all-flash architecture will serve about 80% of its I/O with microsecond latency by not going across the network for reads. When compared to millisecond latency for an all-flash array where all I/O reads/writes incur the network latency, this results in 10X lower latency, which is critical for high performance workloads like SQL, server virtualization, VDI, and big data.

Page 1 of 41234