San Jose, CA – APRIL 23, 2013  – Alacritech, the pioneer in NFS storage acceleration, today announced that the University of Portland has deployed Alacritech’s ANX 1500 Network File System (NFS) acceleration appliance to implement a virtual desktop infrastructure (VDI) across their campus network. The ANX 1500 allowed the University to overcome storage IO challenges and deliver increased performance while cutting both short and long-term costs.

Storage can be a costly performance roadblock within VDI environments. The University of Portland, while experiencing initial success in its VDI pilot, knew that to support a University wide roll-out addressing 3,000 users, including support for its newly re-designed, state-of-the-art digital library, required that its storage be better optimized to better deal with read latency. Left unchecked, NFS writes would consume NAS bandwidth and steal from NFS read performance. Despite using higher speed SAS disk drives and Flash as cache in its NAS implementation, read latency was still creeping in.

To address the VDI storage bottleneck, the University elected to deploy a pure SSD array solution. While latencies did fall, the solution was expensive and represented an adjunct solution to its present and valued NAS infrastructure.

Tom Ank, the storage IT engineer for the University, was faced with the need to add more SSD storage to support the wider roll-out when he was presented with the Alacritech solution. He chose Alacritech because he was intrigued by their ability to provide equal or better performance than SSD storage at a fraction of the cost.

“I liked the idea that Alacritech’s ANX platform could directly integrate with our current NAS system, rather than our current configuration, which was two distinct, separately managed storage systems. Furthermore, the read latency for the ANX was actually less than our pure SSD solution and could be purchased at a fraction of the cost. The ANX also enabled us to use less expensive and higher capacity media behind our filers,” said Ank.

“The fact that the ANX 1500 was able to demonstrate less latency than a pure SSD array, with ample headroom to offload more applications, is a testimony to Alacritech’s data acceleration technologies,” said Doug Rainbolt, Alacritech’s VP of Marketing.

By making use of the ANX 1500, the University expects to save approximately $200K over the next year.  The ANX 1500 will also extend the life of the filer, providing even greater savings. In addition, SATA drives can now be effectively integrated into the storage system rather than more costly SAS hard drives. As a result, the University is anticipating that in subsequent fiscal years, they will save around 50% per year off the cost of additional storage.

About University of Portland

The University of Portland has been Oregon’s Catholic University for more than 100 years and is guided by the Congregation of Holy Cross. U.S. News & World Report ranks the institution as one of the top ten regional universities in the American West. It is the only school in Oregon to offer a College of Arts & Sciences, a graduate school, and nationally accredited programs in the schools of business, education, engineering, and nursing.

 About Alacritech

Founded in 1997 by network storage pioneer Larry Boucher and group of respected storage executives, Alacritech sets new records for the delivery of increasing amounts of enterprise cloud, video and other rich application data.  The company’s first NFS Acceleration appliance, the ANX 1500, mitigates Network Attached Storage (NAS) sprawl and dramatically improves the performance of NFS infrastructures, and does so without requiring the replacement of existing ecosystems or the surrender of ownership of mission-critical data.   The ANX 1500 pays for itself faster than any other storage infrastructure solution and provides by far the lowest cost per operations per second (OPS).  The company holds 54 patents and has received $34 million in funding from Benchmark Capital, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures.  Alacritech can be found at www.alacritech.com, on Twitter: @Alacritech, Facebook at www.facebook.com/Alacritech and LinkedIn: www.linkedin.com/company/alacritech.

Follow Us

Twitter: @Alacritech
LinkedIn: www.linkedin.com/company/alacritech

Press/Analysts Contacts

K/F Communications for Alacritech
Julie Karbo: (415) 255-6505 · Julie@kfcomm.com

VDI is being adopted widely by IT organizations because it can reduce costs, improve data security and simplify management. For example, VDI allows organizations to launch a mass rollout verses upgrading each machine individually, and provides easier transition to and management of BYOD scenarios. But with all of its benefits, VDI can be extremely write-intensive and can cause performance problems.

For example, something as simple as viewing a web page on a desktop generates a number of writes to disk. Add to this the creation of temp files, auto saving and extend this by hundreds or thousands of users, and you have a very write-intensive environment, which in turn can become a bottleneck for reads.

During the video I mention how the customer, a university, was able to use the ANX 1500 to sit along-side the NAS filers in order to accomplish a dramatic decrease in read times and reduce the load times of VDI images by almost 70 percent. The University was able to reduce latency with the help of Alacritech and successfully transition to VDI environment and ultimately optimize performance. The ten minute video session demonstrates some key VDI findings, as well as the overall benefits of server consolidation made possible by server virtualization.

You can view the video discussion with Storage Switzerland’s George Crump below. 

Using Network Caching To Solve VDI Storage Problems

Read original article

Read original article

First, while we’re aware of the benefits of server consolidation made possible by server virtualization, it was quite beneficial to see and hear the effects. What had been a relatively large data center consisting of two very large discreet rooms was reduced to one room. And the one room that remained was only about half occupied. Hence, the benefits of realizing an 18:1 server consolidation. The other large room is now being used as an archive. Not an archive for data, but for art of all things. This particular customer is virtualizing just about everything and making use of Cisco B and C series UCS’s to deliver potent compute and network bandwidth.  To compliment UCS there was ample Cisco Layer 2 and 3 switching and routing infrastructure.  Not only was there the realization of reduced capital and space requirements, but the power budget was greatly reduced. In fact, by taking out power hungry servers and replacing them with more energy efficient infrastructure, rebates from the state were earned. When it comes to networking, the customer is making extensive use of 10GbE throughout, delivering it to each building and major department, with GbE to 10Gbe aggregation switches in each location.

Second, there was extensive use of NFS, making it their standard protocol. Why?  Because it’s easy to use and very compatible with their VMWare environment.  Oracle instances are now executed on Virtual Machines and NFS serves as the protocol for storage connectivity.  The customer makes use of both NetApp and another company’s 100% SSD storage.

Third, this customer is making significant investments in VDI. The primary motivations were costs, data security and simplified management. Desktops are provided throughout the campus where users, who are often very mobile, can quickly tap into their data. Despite the lofty objectives, the customer knew that if the user experience wasn’t good, the project would fail. They couldn’t force users to adopt the technology. So they started small and built up and have succeeded in delivering performance that at least for their applications, is comparable to having a computer with ample network bandwidth at the user’s fingertips.

One thing worthy of calling out in this customer’s VDI experience is what they have been doing to achieve performance. One of the key areas they’ve focused on is write performance, and frankly, it’s been a bit of a surprise for them to learn just how write-intensive VDI is. Someone without a lot of VDI experience may pause to wonder why this could be, especially when desktops may be used for mostly reading data as opposed to writing. What they found is that the underlying OS, in this case Windows 7, generates a great deal of writes to disk as users browse content.  Something as simple and germane as viewing a web page on the desktop generates a number of writes to disk. Add to this the creation of temp files, auto saving, and extend this by hundreds to thousands of users, and you can have a very write-intensive environment with user initiated writes at the desktop only compromising a small percent of the total writes. 

What this customer learned was that writes were becoming a bottleneck to reads – all reads. Reads could constitute securing the boot image to reading a user’s preferences. Remember that if the user experience was poor, such as systems taking too long to boot or to download content, the users would stop using it. To get around this, the customer used the very expensive 100% SSD storage solution to provide enough IOPS to mitigate the write penalty and reduce latency associated with reads. On top of that, they were forced to use much more expensive SAS on the NetApp infrastructure. 

The customer ended up putting an ANX 1500 in its infrastructure to sit along-side the NetApp filers. The ANX could service desktops, retrieving stored boot images and other commonly read data, as well as cache NFS metadata. A tremendous load was taken off the NetApp, enabling it to execute writes much more effectively. The customer was also able to afford paying out big money for another pure SSD shelf for the other brand. Using the ANX the customer found that read times decreased dramatically. Loading time of VDI images was reduced by nearly 70% and launching a mechanical engineering package was reduced by nearly 85%. Using the ANX enables this customer to use less expensive media behind its filers while delivering better performance. And by using the ANX in its mix, the VDI experience is enhanced for users and proves that VDI is viable from both a user and IT perspective. Saving 140 watts per desktop when compared to a PC, delivering solid performance, safeguarding data, and simplifying management all make it very worthwhile.

Seeing is believing. This customer has taken virtualization to the next level without compromising services to users.  Actually seeing such a compact data center deliver so much punch was quite satisfying.

I pushed back… just a little. Fortunately, I work with a candid, talented and strongly opinioned design team. They insisted that the Mac issues could be quickly fixed and that an evaluation on the part of the customer could take place rather quickly. Our Director of Software Engineering, Chris Harrer, was personally involved and even help with any coding and debugging for the few changes that had to be made.  Looking back in retrospect, I think Chris became so convinced that this was a worth project because he spent time with Brian Krusic, Logan’s IT lead focused on storage architecture. Brian not only supports Logan, but is a voice to be reckoned with in the field of visual effects IT design and support. Almost every storage administrator in the industry knows Brian and appreciates his insight. Chris came back more than once to simply say, “This guy ‘really’ gets our technology and knows what he’s doing.”

Brian was convinced that coupling the Alacritech ANX 1500 with commodity based file servers would be a winning combination if everything worked as advertised.  Sure, there was some selling on parts that had to be done, but Brian was quick to envision the possibility to generate phenomenal performance at a cost that couldn’t be beat. His passion rubbed off on our team, and soon we all rallied around Brian to prove that his hypothesis was right; it doesn’t take expensive branded NAS to deliver capacity, performance, reliability, availability and power savings.  Brian had been considering purchasing much more expensive BlueArc storage from HDS. That was nipped in the bud after his evaluation. He instantly saved some decent coin for his company while improving performance. As we talked to more customers, we realized that Brian’s vision is not a one-off.  There are many companies looking for creative ways to store, protect, achieve performance goals and scale using inexpensive capacity coupled with performance layer data acceleration.  This coupling is disruptive and is extensible to many other companies in the film and entertainment space, and for that matter, other verticals. 

There is new case study that can be downloaded on the Alacritech website that describes Brian’s challenges and the steps he took to address them. He achieved some great performance at very low costs. The Logan case study can be found at:  http://www.alacritech.com/resources/. Good reading. We are truly delighted to have Brian and Logan as customers. 

San Jose, CA – November 26, 2012  – Alacritech, the pioneer in NFS storage acceleration, today announced that Logan has deployed Alacritech’s ANX 1500 Network File System (NFS) acceleration appliance to deliver productivity increases and dramatic performance acceleration.  By coupling the ANX 1500 with file servers running CentOS, Logan was able to avoid purchasing excessive and expensive network-attached storage (NAS) infrastructure, costing nearly 50% more in capital. Furthermore, the ANX 1500 purchased, functioning as a performance layer, can support additional NAS systems, when they are brought on line.

Using Alacritech’s ANX 1500, Logan was able to achieve dramatic performance gains. Prior to Alacritech, artists could only use 10 of the company’s 45 render nodes at a time without overloading the storage and even with that, renders took as long as two minutes and 38 seconds per frame. After installing the ANX, all 45 render nodes could concurrently render jobs without experiencing a serious degradation in response times.  In fact, overall render time dropped to just 38 seconds per frame. A big bonus realized was that the load on the filer was reduced by 14/15ths because of the ANX’s ability service client NFS requests.  The ANX 1500 is the first and only storage product to accelerate the performance of an existing storage network without requiring an organization to replace its infrastructure or surrender ownership of data. 

Reliability was also becoming an issue as large demands on the server during heavy render periods caused it to lock up that required rebooting the server.  Brian Krusic, a network consultant at Logan, was faced with the challenge of finding a solution that was reliable, scalable, and extremely cost effective.  Before selecting Alacritech, Brian considered installing an industry standard BlueArc NAS solution.  But BlueArc would have been a much more costly installation requiring additional AC, power, and more physical space. Instead, Brian decided to make use of inexpensive but reliable commodity file servers running CentOS (a free operating system based on Red Hat Linux) coupled with the ANX 1500 to both cache and accelerate data movement. 

“We were blown away by the performance gains we saw at the low price point”, said Brian.  “Not only did the render times drop dramatically, but the fact that artists could access data while all 45 render nodes were rendering was fantastic.  It was a major turnaround and has really increased the productivity of our artists.” 

“The Logan use case is particularly interesting in that Brian was able to apply the same level of creativity that his industry is known for to his storage infrastructure,” said Doug Rainbolt, Vice President of Marketing of Alacritech. “By using the ANX 1500 to maximize their existing NAS system performance, they were able to boost NFS reads and metadata operations into overdrive, giving cycles back to NAS so that it can do more. They did this at a much lower cost than what he would have spent in capital and power required of a more expensive NAS solution.” 

A final benefit of using the Alacritech NFS Appliance is that, because it is a non-intrusive way to inexpensively boost performance, it saves Logan the time and cost of data migration.

“Alternate solutions would have required a highly manual effort to move files between servers, a time consuming process,” said Brian.  “With what we’ve done, the process is so much simpler, and for a small company such as ours, that is a big deal.”

About Logan

Logan is a New York and Los Angeles based creative design, production and visual effects company.  With a talented group of VFX supervisors, directors, creative directors, photographers, etc., they have created such productions as Zombie Land, Avengers and Gymkhana.  Logan can be found at www.logan.tv.

About Alacritech

Founded in 1997 by network storage pioneer Larry Boucher and group of respected storage executives, Alacritech sets new records for the delivery of increasing amounts of enterprise cloud, video and other rich application data.  The company’s first NFS Acceleration appliance, the ANX 1500, mitigates Network Attached Storage (NAS) sprawl and dramatically improves the performance of NFS infrastructures, and does so without requiring the replacement of existing ecosystems or the surrender of ownership of mission-critical data.   The ANX 1500 pays for itself faster than any other storage infrastructure solution and provides by far the lowest cost per operations per second (OPS).  The company holds 54 patents and has received $34 million in funding from Benchmark Capital, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures.  Alacritech can be found at www.alacritech.com, on Twitter: @Alacritech and Facebook at www.facebook.com/Alacritech.

Follow Us

Twitter: @Alacritech
LinkedIn: www.linkedin.com/company/alacritech

Press/Analysts Contacts

K/F Communications for Alacritech
Julie Karbo: (415) 255-6505 · Julie@kfcomm.com
Ashley Hennings: (415) 255-6512 · Ashley@kfcomm.com

There are many benefits that storage infrastructure can provide.  Data captured from Alacritech’s statistical capturing tools among customers indicate that a hit rate of 90% to the ANX 1500 isn’t that all uncommon. This means that only 10% of the NFS operations are passed to the backend. These could be for writes, in otherwise read intensive environments or reads on fresh or modified data.

This hit rate results in several benefits for the user. First, lower latency, which can and often does translate into application acceleration.  One of our customers, a large chip designer, indicated that build times decreased by 40% after integrating the ANX 1500. Another customer, a video effects house, reported rendering times being reduced by 81%. These can translate into a top line benefit.

Another benefit, is a boost to NAS scalability. Taking 90% off of the load off a filer provides ample headroom for the filer to scale before additional filers need to be added. Some of our customers state that they are able to postpone buying new filers. They can use what they have today, allowing them to save money and realize the latency benefits described above. Buying filers less frequently also contributes to less effort being applied to data migrations between old and new filers. 

But one of the bottom line benefits that largely goes unnoticed is power savings. What could a company stand to gain in power savings when using the ANX?  It depends. Factors to be considered are the filer models in question, the number and type of disk drives, the efficiency of the cooling system in the data center, and the local costs for power. But a ball park savings, assuming targeted NFS OPS of 60,000, 2TB of file system data, and mid-range filers, could be in the neighborhood of $5,000 per year, per controller pair in power and about $2,000 in power applied to cooling. That’s $7,000 per year in costs that could be avoided when using the ANX 1500. Now consider that the ANX can support more than just the controller pair used in this example. It can offload additional filers and further scale power savings. This adds more to the bottom line.

Customers like the ANX because it generates incremental performance from their existing NAS investment while saving money, but also because it cuts operating expenses such as power, space, and cooling. I realize that past investments in NAS infrastructure might be considered sunk costs and it’s unrealistic to expect customers to turn off what might otherwise be excess when the ANX is introduced. But many customers are planning ahead, especially given growing data requirements, and factoring the ANX 1500 into the equation now with the intent of cost avoidance in the future. All while realizing significant reductions in latency today.

Read original article

As the multi-device environment has evolved, consequences have come to the forefront for the individual and for the enterprise. With a seamless, rich, inviting and natural user experience, the lines between our work and non-work lives are blurred. We’re giving our companies more of our time; we’re on 24 X7, always plugged in. For the enterprise, it means employees have much more of a voice in, and choice in, and an impact on the devices that connect personal and work time. It also means that the demands on enterprise accessibility and data are growing exponentially.  As a result, more enterprises are adopting BYOD policies and modifying their infrastructures to support this complex, multi-platform environment.

IT has always faced several challenges in the world of consumer-driven device adoption; and with the proliferation of BYOD, a whole new set of challenges exist.   The trade off with allowing employees to bring their own devices to work, be it laptop, smart phone or tablet, is to provide the same user experience across all three platforms while ensuring that data is protected, conforms to standards, ‘fits’ within the existing IT infrastructure, and delivers performance acceptable to users – all without breaking the bank.   This has major implications for the storage infrastructure.

The first and most obvious step in building a storage model is enabling access.  Access Control lists need to be strictly maintained and detail who has access to what information as well as what they can do with it today, this hour, or for as long as a specific flag is set.  The implications of this model suggest that data on a disk stays in the confines of the data center.  Second, the environment likely makes use of virtual desktops and tools such as Citrix’ Receiver, which allows users to access their desktop environment via the cloud and other authorized applications from an encrypted connection to their tablets and smart phones. Users maintain passwords for access to the tablet and access to their virtual desktop and application environments. In the event that the tablet is lost or stolen, the data is safe, stored in the data center instead of being gone with the compromised device.  From the user perspective, the visual transition from desktop, to laptop, to tablet is made as seamless as possible because their user experience, virtualized and supported by IT, follows them wherever they go.  Finally, from the user experience standpoint, IT wants to ensure that the virtual environment provides a satisfactory user experience, which ultimately means supporting today’s important applications, including sophisticated analytics and that performance lives up to the user’s expectations.

Two things can negatively impact performance in virtual desktop environments. The first is boot storms. Boot storms introduce latency into the equation as thousands of users turn virtual desktop devices on. The boot images, stored centrally, must be downloaded across the network and users likely will download specific application data.

The second area that affects performance with regards to latency is the request of read data if the data resides on disk drives in the data center. Requests are typically directed to back-end NAS, making use of NFS. When read-data is requested from say a tablet by way of Citrix Receiver, the user’s access is validated by a server that in turn makes requests of the NAS. If the requested data is not on local caches, it must be retrieved from disks and transported out via NFS to the server.  From there the server uses its own application to encrypt and transport the data out via network, local or wide area in order to then broadcast to the user’s device. This process adds latency and can have a negative impact on the user experience. 

The solution to address these performance concerns is a data acceleration device designed to meet the needs of the multi-platform and high demand application environment.  This device both caches and accelerates data transport from a virtual desktop or hosted application to ultimate end-point clients.  Boot images can be cached and downloaded quickly without disk I/O.  They are then served by NFS acceleration, which further reduces latency. Such a cache is largely relevant in environments where many users are accessing the same information from various end points.  As tablets constitute a great device for the reading of data, latency can be dramatically reduced by retrieving data from the appliance.  With this new class of user- not the creator or “writer” of data – the read to write ratio should indeed be very high.

Such acceleration appliances exist today and are used to reduce latency and accelerate throughput for a number of applications.  In fact, it can be thought of as an application accelerator. From the IT perspective, data is kept safe and secure while simultaneously providing a positive performance user experience; users are able to use their devices to view critical data, but also view it more quickly. The added bonus for IT is that less storage infrastructure may be needed to deliver required performance. And since the appliance satisfies a great deal of the NFS requests, the filer is relieved from having to do so. The filer can instead focus on storing and protecting data, and with a little help from the acceleration appliance, its life is extended. The result is less time spent performing expensive data migrations to newer storage. The end users see a performance boost on devices with less storage infrastructure needed.  Not a bad scenario, especially considering that data is secure and compliance requirements met.  The implications are clear—a diversity of devices, sophisticated applications and the standards required of the more complex IT infrastructure demand a storage infrastructure designed to serve many masters and ultimately deliver the business objectives of the enterprise.