Around the Storage Block

Excited about ExDS

Wed, 14 May 2008 10:40:00 GMT

-by Michael J. Callahan

Since this is my first posting here, I thought I should introduce myself. I am the Chief Technologist for the HP StorageWorks Scalable NAS Division. HP is very serious about broadening its offerings in file-based storage, and the creation of this division was a reflection of that. A significant part of the division--though by no means all of it--came to HP through the acquisition about a year ago of PolyServe, a shared data clustering software company where I was CTO and cofounder. HP had long partnered with PolyServe to deliver NAS solutions using HP servers and storage and PolyServe software, so it was natural for us to find a home here.

It's been a tremendously exciting first year: being part of HP has allowed us to deepen our investment in PolyServe's clustering technology dramatically and also to leverage HP's great strengths in servers, blades, storage hardware, and storage and management software. Last week we announced some of the first fruits of the work we've been doing--the HP StorageWorks 9100 Extreme Data System--and so for my first post I'd like to discuss why I'm so excited about this product.

The ExDS solution builds on PolyServe's experience working with customers who are building new business models based on having vast amounts of data, from hundreds of terabytes to multiple petabytes, available on-line at low cost; for example, web digital media and social networking services, and some kinds of scientific and technical applications.

It's worth acknowledging up front that this isn't everyone. We have customers using our software in a wide variety of ways, from financial institutions running mission-critical applications handling millions of transactions a day, through enterprises building scalable, easily-managed NAS services out of industry-standard servers and consolidating their Microsoft SQL Server databases into highly-available, efficient clusters. For these customers, integration with traditional data center storage infrastructure, like our MSA, EVA and XP storage arrays, is key. We've covered those needs for the past few years with several offerings: the HP StorageWorks Enterprise File Services Clustered Gateway, which includes everything you need to build scalable, highly-available file services on top of traditional enterprise storage; the HP StorageWorks EVA-FS, which adds EVA storage to the EFS-CG; and finally various software-only options, where you add your own servers and storage. ExDS is not a replacement for these products.

In our experience, however, petabyte-scale deployments have a distinct set of requirements, and that's what ExDS is designed to address. Typically, we've found that customers in this space are very concerned about issues of cost, physical density, power efficiency, and manageability/operational efficiency:

- If you're buying such large amounts of storage, obviously cost per gigabyte becomes a major concern. One way of looking at this is: what solution delivers the thousands of drives you'll need for petabytes of data with as little cost overhead per drive as possible?

- Besides cost efficiency, customers are looking for solutions that are physically dense and power-efficient, to make economical use of data center space, power and cooling. Again, you can view this as asking, per drive, how much physical space and power overhead does the system impose?

- Whereas in enterprise environments, it's typical to have storage administrators' responsibilities measured in tens or hundreds of terabytes, in petabyte-scale deployments this would clearly be uneconomical. It must be possible for an individual administrator to handle petabytes of storage.

The ExDS contains a lot of new engineering to provide a complete solution addressing these concerns, building on our experience working with customers who have constructed petabyte-scale environments using PolyServe software. These customers have typically combined PolyServe software with storage hardware of their own choice, in many cases making a significant investment in custom integration to put the whole solution together.

To achieve excellent cost-, space- and power-efficiency, ExDS is based on HP c-Class BladeSystem blades and new storage hardware, different from anything in the existing StorageWorks product portfolio, which we're calling 'storage blocks'.

- Blades run storage services for the outside world, and an ExDS can be configured with anywhere from 4 to 16 blades, based on the level of application performance required.

- Storage blocks contain disks in a very space- and power-efficient package. Each block consists of 82 drives in 7 rack units of space, for a density of nearly 12 terabytes/U with the 1TB drives we'll have. An ExDS can contain from 3 to 10 storage blocks, for up to a total of 820 drives.

Note that performance and footprint can be scaled independently--it's perfectly legitimate to have an ExDS system with 4 blades and 10 storage blocks, or 16 blades and 3 storage blocks. This in itself can be a major source of savings, since you don't have to pay for (and power, and cool) CPUs you don't need.

To address the operational issues entailed by operating at such a large scale, ExDS contains new appliance management software that permits the entire system to be provisioned, configured, monitored and optimized from a single console. Blades and storage blocks can easily be added to an ExDS system using the console, and any failures (drives in particular, in a system of this scale, are bound to fail) are surfaced to the administrator with direct information about what FRU to replace to solve the problem.

Of course, ExDS also includes the PolyServe software, which includes a cluster file system allowing all blades to access all storage in the system simultaneously, provides complete high-availability, and allows standard software--even the customer's own application--to run on the blades within the ExDS system .

I have a lot more to say about all of these elements in future posts, but for now let me just close by saying again how pleased I am to be able to start talking, finally, about some of what we've been working on. From a PolyServe perspective, ExDS does a great job providing a complete, simple, out-of-the-box solution to a big challenge--whereas, in the past, we've seen customers have to invest a significant effort to build similar environments themselves. It also shows why being a part of HP is such an advantage. I mentioned above that the StorageWorks Scalable NAS division, although it includes PolyServe, is far more than just PolyServe. In fact, the ExDS solution has been built principally by a part of our team that was not formerly from PolyServe, and it includes a lot of work, and indeed IP, from other parts of HP. I think it's a great start, and I expect to have a lot more to say about it, and other things we're doing, in future.

I'd love to hear your feedback on ExDS or other aspects of our technology!

Michael J. Callahan

SWD NAS Division CT

Green Storage #3 – Green Storage and SNIA Unplugged Fest

Mon, 28 Apr 2008 09:24:00 GMT

Green Storage #3 – Green Storage and SNIA Unplugged Fest

- by David Garrels

You may have recently seen a recent announcement from SNIA that they were hosting an “unplugged” fest at their facility in Colorado Springs April 22- 23. HP and several other vendors came together to collect the baseline data needed to formulate the standards metrics for power consumption of storage products.

The first step in creating industry standard metrics for storage power consumption is understanding what to measure and how to measure it. And of course, having all vendors measure the same things the same ways. So SNIA hosted an event where the technical teams several vendors could come to the SNIA site and put their systems through the paces.

Each vendor brought a few of their key products (HP brought both disk arrays and tape libraries) and hooked them up to monitored Power Distribution Units (PDU’s) which provided real-time data of the power usage. Systems were tested both in Idle mode and in Peak Performance mode. The data collected from the PDU will be analyzed and used for the next step, which is to determine what common measures can be set by capacity, class of array/library, etc.

One interesting side note – as you might expect when this many products from different vendors are brought together quickly, there were a few minor glitches. Power connections, network configurations and other small issues would come up during the testing. But in the spirit of industry collaboration (and the importance of the work around power metrics & usage), everyone pulled together to get the problems resolved -- good collaboration amongst the different vendors.

We’re putting a lot of work into minimizing the power our storage systems use, and like to hear your feedback on the value of power management and metrics in your data center.

Web 2.0 and Cloud Storage

Thu, 10 Apr 2008 09:23:00 GMT

-by Ian Duncan

There’s a certain irony in writing a blog on what some people are calling ‘cloud storage’ whilst sitting on an Airbus at 34,000 feet looking down on a fluffy white vista… After spending the start of the week at SNW talking to journalists from EMEA and the US there’s clearly a lot of interest in how enterprises are dealing with explosive amounts of data. We have the privilege of talking to Snapfish on a regular basis so we understand how they plan and manage an infrastructure that needs to deal with extreme growth. Snapfish is adding a million customers a month but still manages to have a customer base that has a 90% recommendation rate (as in 90% of them have already recommended Snapfish) – that’s pretty darned impressive.

Our interactions with Web 2.0 customers (fundamentally anyone who delivers content or services via the web) point to three fundamental infrastructure tenets;

1 - The infrastructure has to scale – you have to be able to react to performance or capacity changes very rapidly.

2 - The infrastructure has to be able to be managed holistically – deploying 10PB systems is no mean feat – you need to be able to deploy and live with a system so easily that you’re able to plan for PB’s / Administrator.

3 – The economics of the infrastructure have to align with the business model. Can you imagine how fast YouTube or MySpace would have grown if you had to pay $10 a month (or even $1 a month) to access them? Free works (because you make the money up elsewhere) but you need very, very cheap storage to make the model work.

Here’s the kicker – you need all three, without exception, to make this work for these kinds of customer – scalable and easy to manage at $15 / GB doesn’t cut it and nor does a PB at $3/GB if you need an army of administrators to keep the system up and running…

There’s an opportunity here for system vendors to capitalize on this need – stay tuned for more from HP.

Ian Duncan

Director of Marketing, Scalable NAS

The Right Write Stuff

Tue, 08 Apr 2008 08:19:00 GMT

-by Warren Smith

The Storage Networking Industry Association Executive Director, Leo Leger, announced last week that the most recent individual to gain elite recognition in the storage association’s Achievement Hall of Fame is HP’s John Wilkes. John is just the 2^nd storage professional to be so honored. And we at HP are obviously proud of John and his many contributions to the storage industry and to SNIA. John Wilkes’ contributions included his key role in formulating the Shared Storage Model, long before many people shared storage amongst hosts, and preserving with the SMI-S management interface standard.

With apologies to Humphrey Bogart’s for hatcheting one of his famous line, “Of all the gin joints, and all the plugfests, in all the towns where SNIA assembled storage lab rats, John Wilkes walked in one day and immediately raised the reputation of the joint.”

John is an HP Fellow at HP Labs today, continuing to deliver the best storage engineering on the face of the planet. We salute John and Roger Reich, formerly of many organizations and now President and CEO of his own company, Olocity, as the exhaustive list of honorees to the "Hall of Fame". (No, it’s not a misprint. There are not yet any individuals from three letter companies in the Hall, or from that company with the South Park-esque logo (like maybe an upside down U). What can we say? The competition to get into the Hall is tough! This is one award that no money can buy. Sorry Joe!)

Link to SNIA’s Hall of Fame page: http://www.snia.org/about/profiles

RSA2008 Conference

Tue, 08 Apr 2008 08:13:00 GMT

- by Carlos Martinez

With 17,000 attendees and over 350 exhibitors the RSA2008 Conference can be an intimidating experience for IT storage professionals who are investigating privacy solutions for data-at-rest. HP is a comforting and familiar face for these storage professionals because they know HP is committed to both storage and security with the power of our Secure Advantage portfolio. HP addresses security holistically from the desktop to the data center to protect resources, data and provide validation for audits.

HP StorageWorks has several new Secure Advantage proof points to display at the RSA show including:

A fabric switch designed to offer privacy for legacy tape data.
A simple encryption kit single tape autoloaders and small libraries.
Integration of the Secure Key Manager with the HP Compliance Log Warehouse to extend our value for compliance.
The new HP Storage Security Assessment tool enables customers to gauge their data protection privacy vulnerabilities online and free of charge.

Data-at-rest is a huge privacy risk but HP is definitely there to help with solutions, tools and services which you can see demonstrated here at the RSA Conference.

Implementing an optimized IT infrastructure takes the right combination of technical building blocks

Fri, 04 Apr 2008 09:11:00 GMT

Greetings from the road…. I’ve spent the past couple of weeks on airplanes and have had the opportunity to talk to a large number of midsized customers as well as systems integrators and wanted to take a second to comment on an interesting theme that arose during many of these conversations.

In a sea awash with a dizzying range of interconnect protocol options, disk array technologies, data protection strategies, and other technical intricacies it can be confusing for many IT managers forced to wade through the marketing hype to determine which solution is right for them. Those of us in marketing and sales have been the brunt of cartoons and jokes portraying us as blatant opportunists willing to promise anything during the course of a conversation with a customer but as a former IT Manager I've not completely gone over to the dark side. I’ve noticed many storage vendors hiding behind a one size fits all approach to positioning products to customers and that's no joking matter.

If you are an iSCSI only vendor then voila… iSCSI is the answer to all your prayers. Worried about data protection? Is it any surprise that the up and coming disk based backup only company is going to tell you that tape is old news. The truth of the matter however is that the IT landscape is complex and there is no single technology solution that will be a panacea for customer problems.

Instead I would posit that as companies grow and evolve that their IT infrastructure needs to reflect the best mix of these technologies to answer their unique business problems at a given point in time… it is the mix of these storage technologies integrated with servers and business applications which can truly be called a ‘solution’.

The answer starts at the software layer with a business application which in turn lives on a server that is part of a broader infrastructure which can be made more efficient and better protected by bringing in the elements of storage consolidation and multi-tier backup that are right sized for you.

Your business needs a strong partner who has deep expertise and experience with all of these technical elements and who can bridge the gap between explosive data growth and an optimized end-to-end IT infrastructure.

Call me biased (I am, after all, in marketing) but I’m not making any promises that can’t be backed up by facts. The fact is that HP offers a choice of flexible change-ready technologies to optimize an IT infrastructure made up of clients, servers, applications, networks, and storage and it is the integration of these technologies that ultimately delivers solution value.

My $0.02…. or maybe $0.04

Green Storage #2 – Minimizing Environmental Impact

Tue, 01 Apr 2008 16:43:00 GMT

- by David Garrels

Last time in the Green Storage blog we talked about ways to minimize the power requirements of storage and the three tools you can use today to save power. Now, let’s look at some of the things we’re doing at HP to minimize the environmental impact of our products themselves.

We’ve got Product Stewardship teams from each of our product groups looking at many different environmental aspects of our storage products. The main areas are:

Materials Innovation:

Materials innovation at HP means reducing the environmental impact of materials we use for our storage products. We’re looking at three main areas:

1. Materials substitution and elimination. There are still some materials used in computer manufacturing that not very recyclable, including Brominated Flame Retardants (BFRs) and PVCs. We have plans to replace these as soon as possible across the product line.

2. Reduction of materials quantity. We continually look at ways to reduce the amount of material it takes to produce, pack, and ship our products.

3. Innovative and recycled materials. HP is constantly increasing the amount of post-consumer recycled content used in packaging materials and other product design elements.

Product Recycleability & Manufacturing methods:

Our product stewardship teams are also looking at every low-level part in every product with a goal to drive to zero waste, meaning product can be disassembled to the point where every element can be isolated and recycled.

We also take great care that products destined for recycling under our control really gets recycled. There have been some recent press articles about consumer recycling programs where the material just end up in landfills and we strive to make sure that doesn’t happen.

Product End of Life:

HP offers several choices to manage aging or unwanted computing equipment while also minimizing the impact on the environment. Customers can choose to trade in, return for cash, recycle, or donate their equipment.

Low-impact Distribution methods:

We’re even looking at the ways we ship our product around the world – at how we can achieve our business goals of customer delivery times, and at the same time minimize the carbon-dioxide per unit generated in the shipping process. We do this by balancing the mix of air, sea, and ground methods to optimize delivery and minimize CO2 impact.

So – we’re doing a lot at the product level to make sure our storage products have minimal environmental impact. We'd like to hear your thoughts on environmental impact with your computing equipment & data center operations.

Thin Provisioning: Why was a Leader a Laggard?

Fri, 28 Mar 2008 01:00:00 GMT

- by Craig Simpson

Recently a competitor, EMC, announced that they are one of the last major block storage vendors to join the Thin Provisioning party. That’s not the way they said it, but if you’re keeping score that’s the way it was. Amazingly enough EMC is often considered a leader in the storage industry. So how did a storage “leader” miss such an important development?

The blogosphere provides some good clues. In December of 2006 EMC’s Chuck Hollis blogged, “Don’t Write A Check You Can’t Cash.”

In that blog Chuck decides that the reason thin provisioning is interesting is because, “storage management discipline isn't so great in many shops.” In fact his stated opinion is that, “if reasonable storage management policies were in place, there wouldn't be quite so much interest.” Apparently most of our storage management processes weren’t measuring up to Chuck’s standards.

Now I don’t believe that Chuck and the others at EMC are dumb, nor ordinarily quite so rude. But I do believe they start and end every day as first a storage vendor, then owners of some other IT related businesses. At HP we start every day as a complete IT solution provider. Yes, we’re number 1 in disk storage. But we’re also leaders in servers, networks, IT management, IT services, PC’s, etc. We see the whole picture. We know that management costs are rapidly becoming the dominant cost in IT. Therefore our products and services have to help CIO’s control those costs.

Because we think that way we recognize every app involves an app administrator, a server administrator, a storage administrator, and often more. Those people can’t be meeting every day to make sure every app has all the GB’s it needs. They don’t have time. That’s why technologies like thin provisioning are so important. It lets you keep the apps running without excessive over-provisioning or imposing yet another time consuming process on a group of already overloaded IT personnel.

At HP our people have a great depth of storage expertise. But they also have a breadth of IT expertise through which we recognize the importance of IT administration costs. That’s why you see key technologies like thin provisioning and easy to use storage management products like Storage Essentials from HP. We know that you need a bullet proof IT infrastructure that scales capacity, not headcount. Give us a call to learn how HP can make it happen for you.

Craig

VDI - a hot technology topic at VMworld Europe

Fri, 28 Mar 2008 00:55:00 GMT

- by Ian Selway

It’s been a few weeks now since VMworld Europe and I have to say, that as shows go, it may have been one of the most tiring yet rewarding shows I've recently attended. HP's booth was extremely busy for each of the three days, maybe it was the magician, maybe it was the impressive displays of hardware or maybe it was because we had our new EVA4400 an MSA2000 arrays on display for their first public outings. It really was a great show and a great platform to launch our new products.

But that's an aside; what I really wanted to comment on was what seemed to be one of the two hottest technology topics for customers visiting VMworld Europe, namely Virtual Desktop Infrastructure (VDI). The VDI discussions certainly opened my eyes to what is going to be a really exciting topic over the next few years. That said, what really raised my attention was the various messaging out there on this topic from some of the other storage vendors. I recently read a really interesting blog around storage for VDI and it got me thinking that maybe, just maybe some of the storage vendors out there aren't being quite as upfront with customers about what’s required in selecting the storage requirements for a successful VDI deployment as they might be.

From an HP perspective one of the strengths we have, over a pure play storage vendor, is the ability to scope, design and implement a total end to end solution from the thin client, through to the server and associated storage. With one of the broadest portfolios of any storage vendor, we’re able to make recommendations based on expected workloads, customer technology preferences and future scalability. There’s certainly a need to carefully evaluate how you connect your storage. iSCSI, Fibre Channel and NAS are all choices. A lot of customers are still in the proof of concept stage, but already we’ve seen instances where, depending on customer needs, applications deployed and performance requirements, one offering may be more suitable than the other. Technologies such as de-duplication and Thin Provisioning do have a role to play, but as pointed out in the vinternals article, there are upsides and downsides to these technologies, beyond what you may see presented by vendor videos on YouTube.

We’re certainly working hard at the moment to take this knowledge, our experience with our customers, to develop the complete solution and determine the right server, storage and thin client technology based on actual customer needs. This information will help drive product requirements, as well as enabling us to publish more material to help customers make informed choices across all aspects of their VDI deployments.

One thing is clear; VDI is going to drive the demand for a lot of storage, so it's no surprise that there are storage vendors out there touting their wares for VDI. It's my belief however, that when considering VDI, customers should carefully consider the total cost of the infrastructure including the storage solutions proposed. They need to figure out what's right for their business and to weigh up these costs versus alternatives available, as well as the performance of the solution

We've all seen this before, carefully crafted vendor markitecture pitches and neat storage demos that make it all seem so easy to figure out and deploy. I saw a recent Dilbert cartoon where the manager read a whitepaper and then announced to Dilbert that he’d done his bit, why was the rest taking so long? Following this approach, customers can find themselves taken down a road that leads to lock-in, less than stellar performance or complex hard to manage installations. VDI is a great solution for specific industries and can radically change the dynamics around the way organizations deploy desktop clients. I think that customers, when selecting a VDI solution need to understand what the total cost of ownership will actually be. It seems to make sense to me, that what's required is the need to work with a vendor with experience and a broad offering across the total solution, not just single components of the solution. PowerPoint, demos and whitepapers, are great tools, but they shouldn't replace knowledge and thorough research. Whichever approach customers decide to take, HP is here to help and advise on a total VDI offering.

Ian

SRM.... Not your father's Storage Resource Management

Wed, 12 Mar 2008 22:03:00 GMT

-by Ian Selway

Isn't it amazing, it seems the IT industry couldn't survive without the use of TLAs (AKA: three letter acronyms). Here at HP, we're renowned for our liberal use of TLAs and I often heard the myth that there's some compendium of them handed out as part of the orientation program to help new employees understand what everyone's discussing in the meetings they attend. Well now not only do we have all those TLAs to contend with, but we also now have to understand the differences the same three letters carry when used by different companies. In HP, and across storage generally, for a long time SRM meant Storage Resource Management. Now those of us working in the area of virtualization are coming to grips with a new meaning.... VMware's SRM or Site Recovery Manager.

I've just returned from VMworld Europe, where there was quite a bit of interest in Site Recovery Manager. Everyone wanted to understand what HP was doing with SRM. When would we support it, would they need new software, new firmware and perhaps most interestingly, would HP's adaptor for SRM be free? Well I spent many hours discussing SRM so I though for those of you not at VMworld you may be interested in understanding what HP is doing with VMware to partner around SRM and what it'll mean for your current StorageWorks deployments.

In calendar Q2 2008, VMware will release their much anticipated SRM product. This is sure to generate a huge amount of interest since for the first time, it will bring a level of automated disaster recovery capability that has so far been absent in a VMware environment. Judging by the level of interest from HP customers at VMworld, IT organizations are going to be spending a lot of time evaluating SRM and working out how to deploy within their businesses. That's a great opportunity for storage . One of the reasons the EVA4400 was receiving such attention at VMworld, was the thinking that existing EVA customers could create a low cost disaster recovery site by installing an EVA4400 and using HP Continuous Access software and SRM to link back to their production site.

So why was there so much interest? Partly because to date HP has been fairly quiet about our plans for SRM. Typically we don't tend to pre-announce product until it's actually available to ship. But with so much confusion around about HP's plans for SRM support, it's time to set the record straight. HP will fully support VMware's SRM technology and in fact we'll have the HP Continuous Access adaptor available soon or very soon after the official VMware launch of Site Recovery Manager. Why not immediately, well there's the small task of making sure the complete solution actually functions as it should. We take certification and solution validation very seriously at HP and recognising that data is the lifeblood of the business, we'd rather take a few weeks more to fully test the solution, than release a solution that could put customers data at risk. The good news is that when we introduce our solution you'll be able to have our new EVA4400 replicate data with existing EVA models, a great investment protection story, you'll also be able to use all the functionality of the HP Continuous Access software coupled with the benefits of VMware's SRM solution. Together a great value proposition.

So is there any bad news? Well I guess only for everyone who asked me if the product would be free of charge. My answer is, like everything else in life, nothing comes for free. Pricing for this new functionality is yet to be finalized but when you consider you'll soon have the ability to protect your VMware infrastructure against potential data loss in the event of a disaster, then the combination of HP StorageWorks, HP Continuous Access software and VMware's SRM software will deliver what we believe will be a compelling return for the business. Well worth the investment required.

Kind regards,

Ian Selway

WW Solutions Marketing Manager

Hewlett-Packard StorageWorks

Iomega isn't the only one rejecting EMC's offer

Tue, 11 Mar 2008 22:31:00 GMT

- by Patrick Eitenbichler

Yesterday I noticed a couple of articles in the press re: EMC's proposed acquisition of Iomega -- which has since been rejected by Iomega's board of directors. According to various publications, this acquisition is consistent with EMC's strategy for serving the consumer and small business markets.

This isn't the first time EMC has tried to target their products and solutions at smaller businesses:

In 2003 EMC tried to sell Windows-powered NAS products to smaller businesses -- but then ceased delivery of the NetWin 110 and 200 boxes two years later and admitted that it had a tough time competing with volume server makers in a low-end NAS market. (See "EMC axes Windows NAS")
In 2006, then, EMC launched its Insignia line of SMB products, only to shut it down a year later (accompanied by 40 layoffs) -- again because small and medium businesses preferred other vendors' solutions. (See "EMC restructures SMB storage division")

So why would EMC be successful this time around with Iomega? EMC is known for complex, high-end storage systems -- which is not what SMB customers are looking for.

Small and medium businesses want simple, affordable and reliable storage solutions -- such as...

The HP StorageWorks MSA disk arrays that offer the same management tools as storage inside their ProLiant servers.

The HP StorageWorks All-in-One Storage System that offers application and file storage with data protection "all in one" -- with an intuitive user interface that only requires 5 clicks to move the data of an entire Exchange server to the All-in-One.

The HP ProLiant Storage Server offering feature-rich NAS on industry-standard ProLiants.

The HP StorageWorks D2D Backup System delivering cost-effective, disk-based data protection to smaller businesses.

The list goes on. Simple, affordable and reliable storage solutions -- specifically designed for SMB customers -- something EMC just doesn't offer.

Unencrypted Tape Creates Security Vulnerabilities

Tue, 11 Mar 2008 08:33:00 GMT

-by Carlos Martinez

One of the top storage security vulnerabilities for enterprises today is unencrypted tape. Most enterprises store tape cartridges off the premises as protection against site disaster. This is a good thing. But the unaccounted for cartridge vulnerability arises during transportation or at a 3rd party storage facility. Considering how much sensitive data can reside on a tape and the volume of cartridges handled, it is only a matter of time before some confidential data has unauthorized exposure. Regulations such as CA SB1386 require public disclosure when unencrypted data is lost or stolen. The majority of the states in the U.S. have similar laws. Even international companies doing business in the U.S. need to heed these laws.

In the 2007 the Ponemon Institute study found that only 11% were encrypting tape and it was single digit prior to that. One can assume that most of this tape encryption was software based. Tape encryption is a much more viable solution today because with embedded native hardware encryption, performance is not compromised and some suppliers including HP include encryption in the drive price. Actually the encryption is the easy part of the equation. What requires serious consideration is the key management system because the volume of the keys will multiple over time and data-at-rest keys can live for many years. Enterprise caliber key management systems addressing tape should integrate with LTO4 and be very automated, secure and redundant. Native tape encryption with solid key management will become standard practice in the enterprise in the not too distant future, and then we’ll see SMBs following right behind. Prevention of a breach is much less costly than addressing it after the fact.

“Do something with the tapes!” says Joe Tucci

Wed, 05 Mar 2008 09:27:00 GMT

-by Warren Smith

There is a hilarious new video on YouTube. Strangely, the video was produced by the EMC Corporation to suggest that Joe Tucci ordered employees there to get rid of tape, because the company is backing up to disk (one hopes not exclusively). Titled: “Bob and Joe: Fun with Tape” the premise of the video is however out of sync with Mark Lewis, EMC’s President Content Management and Archiving Division.

While the Tucci-obedient EMC employees in this video immediately begin to cut their tape into streamers and tie them to fans and jump and dance in front of the undulating tape snips, Mark Lewis has not yet amended his posted blog prediction dated August 5, 2007 that, “Here are my top 5 inflection points for storage technology in the next 3-5 years.

1. Offline Storage (Tape) becomes extinct for most uses…”

So who is right on the timing for “Tape is dead” reprisal, #12,435?

Joe – Now

Mark – in 3-5 years from August 2007 (making that 2010 – 2012)

Survey says: Neither of these can touch Nostradamus.

No doubt, either of these executives would love for you to believe that there is no use left for tape, except to create tacky air mobiles. But the reality of the marketplace shows proof positive that people still trust tape for the backups and archiving of much of their data. Fred Moore, well regarded storage analyst and President of Horison Information Strategies wrote in his February 2008 industry perspective “The tape market (Libraries, media and drives) is still a ~$4.2B industry in 2008. Tape cartridge capacity exceeds the capacity of hard disk drives and the gap will continue to widen as long as there is sufficient interest and commitment from remaining tape vendors.” Perhaps there is the thinking behind EMC’s new YouTube marketing: Convince tape storage vendors to give up on a $4.2B tape storage industry and EMC will enjoy migrating all of your data from tape to disk.

Well, HP StorageWorks is not buying into that one, Joe and Mark. Further, we at HP agree with the many people like Fred Moore, who see continuing value in tape storage, in increasingly diverse applications. Although tape’s traditional role as a backup medium has admittedly eroded from its earlier dominance, tape storage technology remains vital for archiving large amounts of less critical information that can be stored safely in case of future need. With email, and now even instant messaging, falling within the domain of compliance legislation, the rate of digital data growth is accelerating faster than ever – and with it demand for tape. As storage capacity continues its relentless growth, customers nearly everywhere are challenged to come up with the funding for the storage infrastructure to maintain this data. Even the wealthiest companies will struggle to justify the cost of preserving their entire content store on expensive disk arrays, even with lower cost disk drives.

At the end of the day, it’s pointless to split tape and disk into separate advocacy groups. HP StorageWorks holds that indeed, it is precisely the differences between tape and disk storage that makes them so effective in partnership. And so, is it just us or does EMC going on and on about the perfection of disk for this purpose and of disk for that purpose remind you of the person that only had a hammer and saw the entire world resembling a nail?

Sorry to hear about the EMC employees wasting their productive time on this video. Tape is not yet dead (footnoted source: Mark Lewis “I said ‘in the next 3-5 years’ “). More data today resides on tape than on disk. And this reality is likely to remain this way for longer than 3-5 years. But thanks EMC, for the video. And did everyone get Tucci’s memo? “Do something with the tapes!”. Thanks Joe! We certainly are.

FCoE…On your mark…get set…Go!

Tue, 04 Mar 2008 09:41:00 GMT

-by Sean Fitzpatrick

The industry is buzzing about Fibre Channel over Ethernet (FCoE) and what its impact may or may not have on iSCSI and network convergence. Initially there were some skeptics in the press like the story from Howard Marks FCoE: The Latest Standard We Don’t Need. But now it seems the tides are turning as you can read about some of the activity on the FCoE website. Putting initial total cost of ownership (TCO) discussions aside; the technology has merit and worth looking at as a catalyst to network convergence and maybe some day replacing iSCSI.

FCoE is an emerging technology that isn’t new but rather revitalized from the work being done in the IEEE (Ethernet) and T11 / Fibre Channel Industry Association. Two improvements have been made to Ethernet include; 1) lossless Ethernet and 2) pause-base flow control. In the late 90’s these two enhancements were not available, hence why we ended up with iSCSI.

FCoE encapsulates Fibre Channel over an Ethernet transport combining the two technologies. In a real application environment, it makes good sense to leverage one common FCoE connection between the server and network enabling new networking options. The good news is FCoE will utilize the same common driver stacks, cabling and management applications you’re already used to dealing with today.

On the other side of the story, iSCSI requires the use of TCP/IP and as we all know has the tendency to drop data packets, unlike FC. Plus, you will take a performance through put hit because of TCP overhead. If you’re deploying iSCSI today you know it requires its own software and management tools separate from networking apps. Because of these characteristics, iSCSI is not suitable for enterprise data centers; rather it is relegated to non-SAN remote storage, stranded storage (i.e. DAS) or small-medium business segments.

It’s anyone’s guess how long it will take for the adoption to take hole, we believe the market will eventually decide the fate of iSCSI or success of FCoE, with a little help of course. If we’ve learned anything from past experience, two things need to happen in order for FCoE to become viable alternative for iSCSI; 1) near cost parity and 2) sustained growth adoption. The premise of converged networks is to help drive out some implementation costs, examples; less adapters, optics, cables will help reduce overall power consumption. This all translates into a much greener environment. HP believes for the foreseeable future (1-3yrs) these two protocols will co-exist, and in or around 2010-11 we should start to see one of them gain more ground over the other.

By comparison:

FCoE will scale, where iSCSI is stuck.

FCoE will be lossless, where iSCSI is not.

FCoE will leverage common management, where iSCSI will always be unique.

FCoE performance is suitable for enterprise data centers, where iSCSI is not.

HP likes the new possibilities FCoE offers to enable network convergence. FCoE has potential to scale in most IT environment from small-medium-enterprise utilizing the same physical Ethernet layer and common management applications in use today. It’s too early to tell, but iSCSI and FCoE will co-exist initially. Once near cost parity is reached…will there be much reason to continue iSCSI?

No, this isn’t your typical Pepsi taste test challenge

Mon, 03 Mar 2008 09:06:00 GMT

-- Buzz Kaas, director of IT, Pattillo Construction

I participated in a HP vs. EMC “taste test” challenge at the HP facility in Houston this past January where we compared the usability of the EVA4400 and the EMC CX3-10.

Working side by side with a SE for each product, we ran through a typical setup of: creating disk groups, selecting RAID levels, creating LUNS, assigning to hosts, several types of virtual cloning routines and tearing it all back out.

Without a doubt, the EVA4400 was easier to use and each task was accomplished in a fraction of the time compared to the EMC CX3-10.

But, more important than ease-of-use is the unique architecture of the EVA. For example, when making a disk group, the controller can auto select the drives for the group. All RAID types (0, 1, 5) are spanned across all spindles in the group. No more hot spots and maximum IOPS for all RAID types. All LUNS are shared across all drives in the group, so all spare disk capacity is pooled together. This makes for a very efficient use of disc space, because you don’t have to have spare capacity designated for each LUN. Expanding a LUN was as simple as typing in the number of GB you want to expand or contract to.

Finally, why risk your data, when you can work on an offline copy in just two minutes? Making clones requires just a few mouse clicks and I can definitely see how we would use cloning for all of our test environments when it’s this easy.

There were three flavors of cloning for the EVA4400:

1) Vsnap: Instantaneous copy using pointers then deltas are accumulated with each page change.

2) SnapClone: Instantaneous copy, but allocating an identical LUN size. SnapClone starts out like a Vsnap, but then records blocks over time to create an exact copy.

3) Mirror Clone: Dynamic copy for DR, takes time for availability, but clone can be fractured, used then reattached.

While EMC does have a lot of granularity, the EVA was far more user-friendly. It enabled me to be more proficient at network storage operations, which we normally farm out, because of staffing constraints. The HP vs. EMC challenge was a great opportunity to really compare the products head-to-head and showed me the value of an easy to use SAN solution.

Reflections on this week's EVA 4400 launch

Fri, 29 Feb 2008 13:33:00 GMT

“Green” Storage – use it to solve real business problems

Fri, 22 Feb 2008 08:58:00 GMT

- by David Garrels

There’s a lot of “greenwashing” going on in the IT world today – vendors talking about their “green IT” strategy in so many different ways it becomes very confusing (buy a pc, plant a tree?). I’ve even seen vendors show up at tradeshows having changed their company logo & booth color to green, just to show how green they really are.

Interesting, but I think “green IT” is first and foremost about getting your business done – and then applying “green” concepts to drive better business outcomes AND better environmental outcomes.

Energy costs are going up (it’s generally accepted that you’ll spend three times as much on powering and cooling your IT hardware over its life as you do acquiring it). Systems are getting more dense and requiring more power. Data centers are “out of power” – i.e. they couldn’t get more power into their data center if they wanted to. And the demands for IT compute and storage resources show no signs of slowing down. These are business problems we can help solve them with green solutions.

After Cooling and Servers, Storage is the #3 power draw in a typical data center. The biggest “culprits” in an array are the spinning drives, whether they’re empty or full, whether they’re idle or busy. And there are three key tools/techniques you can use today to minimize the number of drives in your systems:

#1 – Thin Provisioning/Dynamic Capacity Management. This basically allows you to configure your system to the amount/size of data you’ll need in the future, but only install the number of physical drives that you need now (add more drives when you need them without re-configuring your systems). We’ve seen 40-45% power savings in some cases

#2 – Using “big” drives with storage tiering. A 500GB drive uses about the same power as a 146GB drive (but holds ~70% more). Setting up a disk group with “big” drives to hold data like client pc files can yield a 50% power savings.

#3 – De-dup and Archive. Backing up the same data twice or keeping an “untouched” file spinning forever on disk is a waste of space and therefore power. Using the de-duplication in VLS libraries and archiving un-used data to tape can be between a 40% and 100% savings in power (no on-going power required to keep data on a tape!).

These three techniques are relatively simple things you can do today to save power in your data center. Saving power means either saving money or extending the life of your data center. Those are both better business outcomes that IT can drive. And “green”? Sure, because we all want to save energy and drive better environmental outcomes wherever we can.

Step 1: Admit that IT complexity is making life unmanageable

Thu, 21 Feb 2008 12:02:00 GMT

- by Robin Hensley

We in the IT industry are to blame for building complexity into solutions. There I have said it. Hi, my name is Robin Hensley and I am a recovering complexity junkie. I feel better now as I'm told that taking the first step to recovery is acknowledgement.

I have spent years focusing on individual solution parts (PC's, workstations, servers, storage, networking, software) and looking at each part separately. No longer, the future for both SMB and Enterprise customers is an infrastructure focus with integrated hardware components linked by intelligent, easy to use software. Like anyone who is on a path to recovery I feel the need to tell everyone else about my epiphany. The problem is not everyone else wants to listen. I've discovered I’m not the only complexity junkie. IT vendors and customers alike have engineered complexity into all levels of the infrastructure. And then I found Blades. Blades are not the only way to simplify IT but it is easier for people to grasp the benefits when the whole architecture is designed to eliminate complexity from the start. For smaller IT environments it is easier to think of the entire infrastructure using a "data center in a box" approach.

The HP BladeSystem c3000 enclosure provides an integrated environment for servers, storage and infrastructure. Need 1TB of shared storage? Simply plug in the SB600c All-in-One Storage Blade. No new cabling, no new power infrastructure and it’s as easy to set up and manage as your TiVO. If your email, accounting, customer tracking and even streaming video from your video surveillance system is growing over 1TB, you can expand the storage to 36TB using an external All-in-One Storage System or Modular Storage Array (MSA) with your choice of standard Ethernet cables or faster Fibre Channel connections. Need a more flexible Server and large scale SAN environment? Use HP Virtual Connect to cut down on administration headaches.

On the twelve step IT complexity junkie recovery program, I'm on the path and taking steps to mend my ways. The HP BladeSystem solutions for SMB and Enterprise customers is one step closer to full recovery and I want the world to know because it can lead you on the path to recovery too.