“In today’s enterprise, efficiently managing ever-increasing amounts of data is becoming both a focus of business execution, as well as a serious financial consideration,” said Boucher. “Many enterprise NAS environments are overstretched as a result of the constant influx of unstructured data. I will discuss solutions that both accelerate the performance and extend the life of existing file-based storage systems while eliminating the need to over-invest in expensive enterprise storage systems, thereby significantly simplifying data management and materially reducing capital and operating expenses.” 

Boucher’s presentation will take place at the Westin Hotel, San Francisco, CA on Feb. 14 at 10:00am, PST, as part of the Next Generation Infrastructure track during the two-day conference from Feb. 14-15. 

Pacific Crest is one of the leading investment banks in the technology sector, with a proven track record of success after working exclusively with technology companies for its 21-year existence.  The Summit is structured to provide programs with unique content and interactive formats to foster the highest level of interaction between corporate management, institutional investors, private equity investors and venture capitalists. To provide this unique forum, the Summit is an exclusive, invitation-only event. If you have any questions please contact your Pacific Crest representative.  For more information, please visit www.pacific-crest.com. 

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 Alloy Ventures, Benchmark Capital, Redpoint Ventures, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures. Learn more at www.alacritech.com.

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

For example, in rendering, oil and gas, life sciences, and software engineering environments, performance often becomes a serious issue before running out of storage capacity does. As the loads on storage environments increase, there comes a point of diminishing marginal returns when latency enters the picture and negatively impacts the user experience. Actually, it’s worse than that. It takes longer to execute jobs which can easily translate into poor service levels to customers, decreased revenue, or worse yet, lost customers. This forces administrators to purchase more storage when a filer can no longer deliver performance. While I can imagine that NetApp has a canned answer for this, in the last two months we’ve had two very large NetApp shops tell us that data migrations are extremely painful. Fortunately, there is a happy medium between suffering through degraded performance and NAS sprawl.

In operating environments that are read and/or metadata intensive, we’ve seen a great deal of the load removed from filers when the ANX 1500 is present. This is simply because we reduce the amount of traffic that must be handled by the filer. We’ve seen ANX customers realize a hit rate of over 90%, meaning that less than 10% of the traffic has to go to the filer. And with that the latency associated with this traffic is greatly reduced. One customer reported that software build times decreased by 40%. On top of that, the NFS write and CIFS performance increased substantially. This was made possible because the ANX appliance took the load off the filer, enabling the filer to do more, even for operations that we don’t directly touch. When the filer can do more, it doesn’t need to be refreshed as frequently, and the pain associated with data migrations can be improved significantly. One customer stated that he could now put off buying new NAS by a year, and this after increasing the number of clients directed towards the ANX/NAS combination by over 3X. Even with the increased load, the CPU utilization on the NAS dropped from 70% to 40% during the peaks, and traffic was reduced to the NAS by 50%. The customer said that they expect to increase loads while providing their users with a better experience. And clearly there was headroom on the ANX to offload additional filers. We estimated the net savings associated with using the ANX and forgoing buying more storage to be approximately $250,000 in year one. And when they do buy more storage down the road, it will go further because the ANX will have the bandwidth to offload it too.

We’re seeing some improvement in the economy, but companies are still looking for ways to boost performance while eliminating unnecessary storage expenditures. An approach that delivers something in between pulling your hair out or your equipment out is a safe and sane choice.

Winners will be announced at the Storage Visions Conference award reception on Sunday, Jan. 8, at 6:00 p.m. PST in the exhibit hall of the Riviera Hotel Convention Center in Las Vegas, NV.

“We are honored to be a Storage Visions Visionary Awards finalist,” said Doug Rainbolt, V.P. of Marketing. “It validates our vision of improving the performance of existing enterprise network storage infrastructures while drastically decreasing costs.” 

About Storage Visions Conference

The Eleventh Annual Storage Visions Conference held at the Riviera Hotel Convention Center in Las Vegas, Nevada, January 8 & 9, 2012, features great sessions on media and entertainment technology, consumer electronics, computer applications and all aspects of digital storage with the theme of “Heavy Storage for Thin Clients.”  More information is available on the conference at www.storagevisions.com. 

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 Alloy Ventures, Benchmark Capital, Redpoint Ventures, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures. Learn more at www.alacritech.com.

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

• Data Growth is Inevitable. Whatever happens in 2012 with respect to our worldwide economic situation, data growth is a given. Unstructured data represents by far the most significant management challenge. And most of this data is accessed on a random basis. For enterprises looking to assure decent data access performance this can be an expensive proposition. Survey respondents told us that the challenge is made all the more difficult given the fact that they face flat budgets in 2012. Only 20% indicated that they plan to increase their spending. As vendors, we can help by continuing to provide cheap capacity combined with precise performance boosts when and where needed. This coincides with my belief that we will see a bifurcation in architectures: a capacity layer for storing and a performance layer for prescriptive performance acceleration.
• Targeted Investments. The first area that stood out, not surprisingly was traditional capacity. Investments in SAN and NAS were about equally footed. I was bit surprised that traditional NAS spend (scale-up) was higher than scale-out. This might have more to do with legacy and I’d suspect that we’d find that scale-out NAS is growing faster. The second key area where customers are planning on making investments is in cloud-based storage services. The economic realities associated with managing storage in traditional ways, versus a utility model, are likely shaping this need.
• Use-Cases for Flash Memory. Very few of the respondents stated that Flash would replace traditional disk drives. Most felt that Flash has its place alongside conventional storage as a tier or as a cache. But, and I found this interesting, many of the respondents weren’t exactly sure how it would be used “alongside” traditional storage. In other words, they’re still exploring options. This tells me that many users aren’t totally convinced that using Flash as storage tier (tier0) in a capacity layer controller and/or node alone is optimal. And this would be good news for vendors that can demonstrate that using Flash as a key component of a performance layer architecture provides better performance, while driving costs much lower as well as simplifying management. When asked how Flash is being used today, most of the respondents indicated that it’s used as a cache as opposed to storage. Given the NAS vendor profiles (below), it’s not hard to imagine that much of this constitutes NetApp® Flash Cache cards, which have enjoyed a high attach rate, especially for the FAS 6000.
• NAS Vendor Choices. Not surprisingly NetApp’s dominance is still very visible, with over 70% of respondents stating that NetApp is their primary NAS vendor. However, we’re also seeing customers mix and match vendors. It’s not uncommon to see NetApp and Isilon® gear in the same data center, with Isilon being preferred when scale-out is needed. When we meet with customers many state that they are waiting to see how NetApp performs within a Cluster Model. If this goes well, NetApp might slow down the advancement of Isilon. If it goes poorly, I would expect Isilon to take more market share. Despite this, we shouldn’t assume that traditional scale-up NAS is dead. A small amount of respondents made reference to using BlueArc. Our experience has been that BlueArc customers have tended to standardize on BlueArc alone; we don’t see the mixed environments that we see with NetApp and Isilon. The good news for a company such as Alacritech is that we don’t care what NAS vendor we’re accelerating data for. An open approach provides the freedom to deliver a performance layer architecture that complements a capacity layer—delivering the best of both worlds.

Join the conversation on Twitter by following @alacritech and share your opinions by using @alacritech and the hashtag #alacritech12 in your tweet.

Happy Holidays to you and yours,
Doug Rainbolt

While it’s easy to get comfortable with what we know, we should all be questioning the status quo. Based on your existing storage infrastructure, your resources, the characteristics of your data and the access required, this is a perfect time to take a fresh look at the storage options in the market. It’s key to evaluate how they can increase performance while reducing, or at the very least, maintaining operating costs and capital expenditures as your storage demands accelerate.

What Are The Best Storage Options?

1. Thin-Provisioning. Since being popularized by 3Par, and well before being acquired by Hewlett-Packard, thin-provisioning technology plays an extremely important role in provisioning storage capacity. I remember in the early days of SAN, there were those that postulated that enterprises were actually using more capacity by over-provisioning in anticipation of data growth. With thin provisioning, enterprises can provision for what they need while adding capacity as needed, without the headaches of carving out new LUNs (logical unit numbers). I learned from a 3Par insider that a technologist within the company came up with the idea in response to a challenge to differentiate 3Par, not fully appreciating at the time just how powerful this would become.
2. Object Oriented Storage and REST (Representational State Transfer). Initially, this technology will have a bigger impact from the cloud perspective. Increasingly, enterprises will feel more comfortable storing data in the public cloud. HTTP will likely be the transport, making use of REST as a means to move and store data while providing the rich meta-data descriptions accompanying data. Initially, I see the primary use cases as being across the wide area, but in the long-run don’t be surprised to see it take hold in the data center.
3. WAN Optimization. The efficiencies that can be gained by using WAN optimization products, such as Riverbed Technology, are extreme. Decreasing the amount of traffic sent across the WAN, by both removing duplicate data and compressing the remainder, can result in significant storage savings, decreased latencies, and decreased expense associated with WAN bandwidth. As enterprises find themselves both viewing and creating data throughout the world, WAN optimization is key to enabling users LAN-type access while keeping a lid on storage, networking, and storage expenses.
4. Tiered Storage. As enterprises look to balance cost and performance, the idea of storing data on the best media to match data value and performance expectations makes sense. Data that is very infrequently accessed shouldn’t necessarily be storage on SSDs or higher performance disk-drives. Vendors have introduced storage products that feature automated data placement based upon access patterns. Flash can constitute either a storage tier, often labeled as Tier 0, or as cache. There are benefits to both. Enterprises should understand how data is placed on which media and model data growth to better understand the cost of scaling capacity and performance before making a purchasing decision.
5. Scale-Out NAS. Traditional scale-up NAS as we know it will transition into becoming a smaller and smaller percent of the mid-range and enterprise-range of the NAS market. It’s being replaced by scale-out, which provides enterprises with the ability to add incrementally, in a clustered fashion, both capacity and performance while working under the construct of a Global Name Space. Provisioning of storage can be greatly simplified and it’s not uncommon to find a single storage administrator managing petabytes of data. The total cost of ownership can be greatly reduced.
6. Performance Layer Appliances. Think of appliances that are less concerned about managing NAS based capacity. These store data and are more concerned about moving data with maximum efficiency – i.e., an appliance that not only caches data but accelerates its placement on the wire. It is NAS optimization that gives IOPS back to the NAS, an appliance unburdened with all of the heavy computation duties reserved for the NAS. The result is increased performance and significantly less capital and operating expenses.
7. FCoE. FCoE, or Fiber Channel over Ethernet, helps enterprises extend the reach of Fibre Channel across Ethernet infrastructure. And saving money on infrastructure, including cable and power management, is a really big deal. Ironically, we don’t read much about FCoE anymore. Nor do we hear or read about the raging battles between iSCSI and Fibre Channel. But this doesn’t mean that it’s not going to become important. The winner is Ethernet. Many enterprises openly acknowledge having both ample block and file traffic and are looking for products to handle both. While the growth rate for file based storage is outpacing block storage, don’t expect the need for Fibre Channel or FCoE to go away anytime soon.

Making new choices doesn’t have to put your current infrastructure and data center management at risk and taking a look at some of these options can reap very real financial rewards.

Compared to traditional storage, where data is stored on the same media, tiered storage is significantly more efficient. However, efficiency is a relative term and has to be looked at over time. Tiered storage may be efficient today, but in six months data could have grown substantially and what was once the right size storage may not fit any longer. This forces designers to over-design arrays, plan for the possibility that the tier 0 storage may be inadequate or that data may need to come from disks as opposed to SSDs. The primary role of these systems is to store data – a capacity layer. Once additional features are added, including “tiering” software (policies that dictate what goes where), you have a system that, although better than older designs, could benefit from better efficiency. These systems execute multiple processes, processes that can consume inordinate amount of resources and can actually conflict with each other. The primary role of this capacity layer is to store data, but the challenge is that we’re throwing more and more data into the mix, expecting this layer to do more.

While this capacity layer is adequate, imagine having another layer that sits alongside it and complements it. This layer isn’t encumbered with the overhead and duties of storing data; its role is to accelerate data movement with maximum efficiency – a performance layer. By using such a layer, you can be much more efficient in how much capacity layer infrastructure you use. In fact, you’ll probably need much less of it. This performance layer can help you cost effectively scale storage, while delivering performance. And, the payback can come in two ways: it may actually allow work to be completed faster and hence contribute to revenue generation and two; it can reduce costs, capital, power, space, and cooling. And it does this by taking advantage of the NAS that you already have.

In a symbiotic relationship, the performance layer takes on data movement which otherwise could slow down the capacity layer. This allows the capacity layer to focus on storing additional data. Both layers complement each other in maximizing the NAS storage capacity you currently have, while improving performance – something that even my seventh grade science teacher would appreciate.

Read original article

WHAT: “What Does Tiered Storage Really Do to Performance” at Storage Decisions San Francisco 2011

 WHO: Alacritech V.P. of Marketing, Dough Rainbolt

WHEN: November 8, 2011 at 11:35a.m. – 11:45 a.m. PST

WHERE: Hilton, Union Square. 333 O’Farrell Street, San Francisco, CA 94102

About Storage Decisions San Francisco

Storage Decisions San Francisco, scheduled for Tuesday, November 8, 2011 at the Hilton Union Square, is the only event built exclusively for IT pros focused on the storage, backup and disaster recovery planning challenges that every organization is now facing. The 200 IT pros who claim delegate seats will hear (and actively participate in!) sessions and workshops conducted by the top independent experts in the storage field.

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 Alloy Ventures, Benchmark Capital, Redpoint Ventures, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures. Learn more at www.alacritech.com.

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

The implications on performance from a guest’s perspective can’t be overlooked. How can we ensure that applications are best matched to the right physical storage both today and tomorrow? And, how do we ensure this matching without breaking the bank?

In this translation from the physical to the virtual world, the linkage between an application and its physical storage has obviously changed. In the pre-virtualized world, storage administrators would talk (another abstraction) with the application team when carving out LUNs or file-based storage that provided the capacity and I/O performance needed. Today, it’s more a case of the storage folks providing LUNs or file-based storage to the team managing the virtual infrastructure, not a specific application. There is surely knowledge as to what to what applications demand in capacity and I/Os, but when multiple applications are executed on a physical server, from the storage target’s perspective, data can look far more random than when compared to separate physical servers, each hosting a single application and specific storage provisioned for that application.

The question becomes what is acceptable performance in these highly virtualized environments and how much does it cost to deliver the performance? As data from the storage target’s angle becomes more randomized, there needs to be sufficient IOPS to deliver the data throughput the applications, in aggregate, demand. If the intention in the data center is, like server virtualization, to get the most value from an economic perspective out of storage virtualization, having a performance layer accelerated cache that can generate extremely high IOPS can be an extremely economical solution. Storage arrays can’t match the low $/IOP that a purpose-built appliance can generate. Why? Because a network acceleration appliance, constitutes a single-stack design and is “purpose built” to process NFS requests at the lowest latency possible. This is why the ANX 1500 produces the lowest published ORT in the NFSv3 SPECsfs® 2008 benchmarks.

A network acceleration appliance can take a great deal of the load off of the back-end storage, which means it can perform better. In fact, you may need less storage without compromising performance. Obviously, the higher the hit rate to the ANX 1500, the better the performance. Users having an NFS mix similar to the SPECsfs mix, or even higher percentage of reads, stand to realize significant performance gains and will rely less on the backend NAS to deliver performance. And these performance benefits are realized by getting more out of the NAS investment you’ve already made.

“Recently there’s been a great deal of attention paid to many of the elements of a storage infrastructure,” said Rainbolt. “Some people are advocating spending large amounts of money on flash, but actually the key is to buy a range of products that will net the performance gains you want within the parameters of a reasonable budget. On September 20th, we’ll cut through the marketing hyperbole and talk about evaluating your options – whether it is flash, acceleration appliances, traditional drives or a combination of several alternatives – and getting to a decision that provides the most upside for your company.”
Rainbolt’s presentation is part of the Pavilion Sponsor and Emerging Technology Showcase and is scheduled for 9:55 to 10:05 a.m. EST on Tuesday, September 20th at the Hilton New York, Gramercy Ballroom.

The two day Storage Decisions Conference is the place to find the industry’s best independent experts – as keynote speakers, the leading vendors in the storage space – as valued sponsors, and the most qualified group of storage professionals – as attendees, all under one roof. The carefully selected collection of experts, end-users, and sponsors provides those involved in the evaluating, implementing, and purchasing of storage an unprecedented opportunity to gain insight into the products, policies, and procedures driving storage demands in 2011. For more information, visit http://storagedecisions.techtarget.com/newyork/index.html.

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 Alloy Ventures, Benchmark Capital, Redpoint Ventures, Berkeley International Capital Corporation, Munder Capital Management, Needham Capital Partners, and Quantum Technology Ventures. Learn more at www.alacritech.com.

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