Last week I spoke at the Virtualization Group - Boston event held at Microsoft's Waltham offices. I enjoyed meeting the group very much and let me thank Tim Mangan for letting us participate in that event.

Tim has continued that conversation about AutoVirt file virtualization and Microsoft DFS online at his blog, and it has been interesting to follow along. Now seems like a good time to chime in. There are a couple of items in the comment thread worth addressing – so hopefully I can explain some of the reasons our customers need the capabilities that AutoVirt supplies.

1) What’s the big differentiator to DFS? 

Let me tell you the story of one of our customers.  He had five Windows file servers with 6,000 shares that were used by 2,000 people in his company.  He needed to move it all to a new NetApp without impacting his users.  After installing AutoVirt he created the namespaces in 20 minutes, converted the 2,000 users from direct access to access via the AutoVirt namespaces in 5 minutes, and kicked off migration in one minute.  The next six weeks, AutoVirt on his behalf, migrated all the shares and users on nights and weekends. After watching the rate of progress, he decided to speed it up.  He added 10 servers from a blade farm to the AutoVirt configuration and it automatically redistributed the workload.

His users never discovered what was going on and he took the time originally allocated to work other things.

Now, let’s examine the differences between DFS and AutoVirt and imagine the same story played out with DFS.  

1. DFS namespaces are created manually with a GUI or a script.  Creating a namespace with 2,000+ folders, 2,000+ sets of ACLs, and 2,000 targets that way would take a prohibitively long time.  Testing and debugging the result afterwards would be similarly demanding.  AutoVirt does all that in 20 minutes using its automated file server inventory function.
   
   
2. Changing all the shortcuts and embedded links used by the 2,000 users and get that right would be equally monstrous. Some of our customers have million of embedded links in databases and files making it impossible rather than hard. AutoVirt makes it entirely unnecessary to change any of these by transferring the file server names to its namespaces in the domain controller and renaming the physical filers.
   
   
3. DFS has no data management capabilities.  The DFS manual makes the following suggestion for data migration:
   1. Install the new filer
   2. Create all the 2,000 shares with ACLs on the target
   3. Take the old file servers offline
   4. Copy files and ACLs to the new server with RoboCopy
   5. Edit the 2,000 namespace targets manually in the DFS namespace
   6. Bring the clients back online (and keep your fingers crossed)

The AutoVirt migration policy does all that, no manual labor required, no downtime required.  In AutoVirt, policy and namespaces are automatically kept synchronized.

In DFS, the namespace is the purpose.  In AutoVirt, the namespace enables data management independently from data use.

In fact, there are many differences between AutoVirt and DFS and if you are interested in learning more then you should check out the DFS Tech Brief on AutoVirt's resources page.

2) What about reparse points and MKLINK? Why do you need a broker like AutoVirt instead of just using the built-in reparse points?

Agreed, if a namespace was my only purpose then I would use re-parse points instead of DFS or AutoVirt.  Re-parse points are fine for file system extension or for adding capacity as long as all the users come in the same way and are willing to traverse through them.  File systems that are built this way makes data management harder since re-parse points contain additional physical references.

A global namespace that is delivered separately from the participating file system creates an abstraction that provides stable views for the users while allowing the IT staff full freedom to relocate, replicate, and provision. Let me give a few examples:

1. Data migration. With a namespace, the content of any target can be relocated independently and at any time.  This enables server replacements, consolidation, splits and merges, and tiering.
2. Copy Replication. With a namespace, the “same” data can exist in multiple places.  Clients will use a single URL to get the full list of referral targets and will select the closest copy regardless of where they are at any time.
3. HA Replication. Two copies of the data are maintained, one as primary, the other as a standby.  The namespace is used to switch the clients.  Policies are used to synchronize.

These examples only work as long as the namespace is independent from the file systems.

In summary, AutoVirt automates the management of data and file system resources, the namespace is a feature, not the goal.

Check out what we are doing at http://autovirt.com. 

This is how people lose documents, and how duplicate files proliferate across your network.  This is how backups slow down, and why recovering backed-up files is difficult.  So how do you force users to comply with the organizational system you set up? 

One way is to enable access-based enumeration so that end-users see only the shares that they are allowed to access.  This simplifies their view and may prevent them from creating new folders when they can’t find what they want. 

Another way is to create user- or department-specific views of the shares on your network.  You can do this by using a file virtualization product to virtualize file access.  Then, you can create numerous hierarchical share arrangements, and give users access to their own unique arrangements.  As with access-based enumeration, you can show users select shares, and hide the rest. 

AutoVirt is a drop-in replacement for StorageX, but we were a bit surprised when we learned why our customer had bought StorageX and was now interested in AutoVirt: backup.  AutoVirt is not a backup tool, but it turns out that some impediments to backup strategy may not be addressed by backup tools.

When user shares and file system directories are disorganized, it is difficult to meet backup, security, and retention policies.  Suppose you have different backup requirements for different departments, as our customer did.  If department-owned shares aren’t clearly defined, it is impossible to create department-specific backup policies.  Instead, you have to back up everything to meet the tightest set of regulations.

This isn’t necessarily a problem until backups begin to take too long and your backed-up data takes up too much space.  And if you ever need to recover backed-up data – because isn’t that the point? – it won’t be easy to find the files your users need.

Our customer had a plan: he wanted to create a new directory structure on a new file server, and move shares from the old file server into the new structure.  The problem was that the reorganization would take too much time and create too much end-user disruption.  Enter AutoVirt.

We’re excited to hear how the project progresses.  We’re also wondering how many other people out there are facing backup policies stymied by share organization.   

Alas the darn project had run long once again
And the CTO as always had ordered his men:
“You’ll stay here all night ‘til the problem you fix
Lest when they come in all the users be pissed.”

So the IT staff was pulling yet another all-nighter
And this one now promised to be a nail-biter.

They were moving the firm’s data from a tired old NAS
When someone discovered some shares that they’d passed:
Enormous in size and quite critical to all
The shares?  Very big.  Time to morning?  Very small.

A fix to this problem was hard to conceive
And everyone on staff was past ready to leave.
When suddenly through the doorway Klavs Landberg came bounding
Wearing AutoVirt blue and cheerfully shouting,

“Don’t worry, with AutoVirt your problems will vanish
Our simple global namespace makes it easy to manage.
You’ll move all your data now faster than ever
What you don’t move tonight: move tomorrow – move whenever!

Users access the old server just the same as before
And with AutoVirt in place you’ll benefit still more:
When you move any data from old server to new
AutoVirt automatically patches your users straight through.

Move a share at a time – or three, maybe four
Keep both servers live ‘til you’re done with the chore.”

The room silenced with surprise.  Could it be Klavs was right?
With AutoVirt installed, they drove home for the night.

Next morning IT staff came in rather nervous
Had the AutoVirt bet left users without service?
A quick survey revealed that all had gone well
Data was spread between servers, yet no one could tell.

Do you have a migration?  An upcoming project?
Want a new way to do it, an easier object?
Sign up for a demo or check out our website,
We can’t wait to hear from you

…and to all a good night.

Back in May, Brocade quietly announced that it was terminating its whole StorageX file management product line. The official StorageX End-Of-Life (EOL) letter to its customers is certainly worth reading. I guess the company has to focus on its core business – after all, you’ve probably seen the reports that Brocade has been shopping itself around. 

StorageX was a file virtualization product much like AutoVirt 3.0. There were a bunch of customers using it – and now it is dead. In fact, the damage extends beyond just Brocade StorageX… there are a bunch of impacted products from both Brocade and its resellers:

• Brocade StorageX
• Brocade File Lifecycle Manager (FLM)
• Brocade MyView
• Brocade File Management Engine (FME)
• NetApp Virtual File Manager (VFM) product suite
• IBM StorageX
• Hitachi StorageX

Now if you are a StorageX customer, here’s the crazy part. You won’t be getting any new product enhancements – no features, no upgrades, no new platform support. And in exchange, you have to pay 10% more per year, each year, to maintain your support contract! Thank you sir, may I have another?

We’ve been getting so many calls from StorageX and NetApp VFM customers looking for a replacement, that we decided to create a StorageX Swap-out Program to make it easy for people to transition off their unsupported platform onto something that is actually easier to use, easier to manage, and even more functional.

Here are the details of the program:

• A 50% discount off perpetual software licenses of AutoVirt
• One year of AutoVirt support at no cost
• StorageX import functionality that automatically recreates the entire global namespace with just a couple of clicks – for an incredibly simple transition process

Want to learn more about the program? Find more details on our StorageX Swap-out Program page.

The real point is that we are both rapidly building momentum in the marketplace - file virtualization technologies are alive and well, and the industry is growing.  Customers have by and large completed their server virtualizations and they are now focusing on manageability of their vast file systems. The industry has solved the hardest challenges and it is creating file virtualization products that are cost effective, easy to use, scalable, non-disruptive, and supportable.  The time is now.

Two winning solutions have emerged - ARX is the leading in-band solution and AutoVirt is the leading out-of-band solution.  As they say, a rising tide lifts all boats. The fact that we are communicating is remarkable and that will ultimately benefit everyone.  Management of file data is a huge and pervasive problem that deserve all the help it can get.

File virtualization can be inserted transparently, meaning that all pre-virtualization shortcuts and embedded links continue to work without change.  Failure to provide transparent insertion killed NuView StorageX (acquired by Brocade and pitched as a File Area Network (FAN)).

File virtualization can provide scalable performance.  Limited scalability killed NeoPath and Attune.  EMC/Rainfinity is stubbornly still around, but has not been widely adopted.

File virtualization can offer integrated data management.  Lack of data management functionality, and the manual effort required for namespace creation and insertion, has doomed widespread use of DFS from Microsoft.  If it is free, why not use it?  Why indeed.

File virtualization can offer support for existing IT practices like server-based backup and restore, snapshots, and scripted procedures that work directly with networked file devices.  It is still unclear how important compatibility is to file virtualization adopters.  Despite its incompatibility with existing infrastructure, ARX from F5/Acopia continues to win customers.

The Next Stage in File Virtualization

With sincere apologies to Mark Twain, reports of the death of file virtualization are greatly exaggerated.  File virtualization is alive and well.  The out-of-band solution from AutoVirt passes all four tests and more with flying colors.  With apologies to Larry Niven and Jerry Pournelle: stay tuned, we are making hot fudge on a cosmic scale.

Tiering is all the rage – we’re hearing it everywhere. Here's one good article on tiering posted just yesterday at SearchStorage.com. IT organizations, driven by rapid data growth and the pressure to drive costs down, are trying to figure out how to build storage pools that are high-performing, scalable, yet cost-effective. And what they’ve realized is that this can be achieved by classifying data according to various attributes, and then housing that data on storage tiers specifically designed for those characteristics. 

But inherent in this architecture is this question: How do you get the right data on the right devices without affecting users, apps, and management processes?

The most immediate solution is to purchase a storage appliance that houses different tiers internally – or at least front-ends them – and as a result creates an abstraction layer that hides the varying tiers. This is exactly what the major storage vendors like EMC want to see happen, as well as lots of startups – Avere, for example. 

But I believe the market will reject this type of solution for a number of different reasons.

1. Vendor Lock-In: Storage technologies are changing rapidly, and competition drives innovation up and costs down. It’s important to have choice, flexibility, and access to the best technology that fits your environment. No one wants to be locked into a single vendor and forced to pay a premium for it, especially when there are so many exciting and inventive products that may provide more value.
2. Every Environment is Unique: You need to be able to design tiered configurations that are optimized to your environment – the size of the various tiers, the proximity of data to the apps and users that use it, the specific replication, backup, and management policies of different types of data, and more. There is simply too much complexity for a one-size-fits-all approach to storage architecture.
3. Single Points-of-Failure in I/O: There is something disconcerting about a hardware solution where every bit of data has to go through a single point-of-failure. Especially if that is the gateway to all the unstructured data in the entire organization. 
4. Expect the Unexpected: You buy hardware expecting it to last around 3-5 years, but no matter how well you plan, it’s impossible to really know what your data needs are going to be during that timeframe. That’s why you architect for flexibility, so you can easily swap in new technologies and be responsive. It’s hard to do that inside a box.

What IT professionals really want is to be able to “see inside the box.” They want the control and flexibility of being able to assemble and architect the hardware components to meet their highly specific needs, and they will need scalable solution to manage that environment so that it presents itself to users and applications as a single pool of storage.

Read more about how AutoVirt helps you build out a tiered storage environment from 3rd-party hardware.

Q: Can you explain a bit more on what you mean by adding objects to the DNS config? Perhaps an example?

The AutoVirt global namespace is made up of individual namespaces. The root of each of these namespaces is found by client computers via the DNS entries (a.k.a "objects") that the AutoVirt software creates to describe them. DNS entries are also used to create the highly available namespace service created by individual namespace servers working together. After accessing a namespace and receiving the referral information, clients go back to DNS to locate the filers, again using DNS entries created by AutoVirt.

Q: If your AutoVirt servers become unavailable, does that mean users will not be able to access file shares?

AutoVirt is built for high availability. Two dedicated servers (called ‘nodes’) each host the global namespace, policy engine, and data engine. These servers are identical. Should the primary node become unavailable, users would be immediately redirected to access their data through the other node. There is no magic, however: if both servers become unavailable, clients will not be able to access shares.

Thanks for your thoughts, and keep ‘em coming. We hope you’ll view the recorded webinar and contact us with further inquiries!

Here are some additional thoughts on the state of the storage industry as seen during the Storage Networking World (SNW) 2009 event in Phoenix, AZ:

Fewer vendors, but end-users aren’t slowing down

The economy has clearly impacted companies’ abilities to sponsor and even attend shows like this. Anecdotally we heard that the show floor was only about a third of the size that it was last year. Cisco & Brocade were obvious absences. However, end-user attendance didn’t change much from last year – giving the end-users a much greater advantage in numbers over vendors. The result? We really got to connect with people from a wide variety of companies and discuss a range of business problems. Sure – budgets are tight – but it felt as if most had cut what they were going to cut, and now the challenge was improving the efficiency of the existing staff by operating smarter, faster, and in a more automated fashion. 

The need for speed

The hottest topics seemed to be related to speed. There was one vendor – I wish I remembered the name – who had a display of about 8 HDTVs, each screen split into small individual cells that each played an individual movie. In front of the display was a plastic-encased piece of electronic equipment and a the tag line that read was something like “1200 DVDs streamed from 1 drive.” Other vendors big on de-duplication for the purpose of speed included Whiptail and Exagrid – there were many more. After hearing similar needs echoed in the hallways, what I take away is that the virtualization process has created new storage bottlenecks, so that’s what people need to solve next. I feel like this is a good subject for a future blog posting.

The Need to Manage Unstructured Data

Our booth was packed and we were thrilled with the level of interest and activity. But why was that the case? I think the momentum can be attributed to three things:

1. The session on File Virtualization, led by AutoVirt customer Mike Williams from Five Point Capital. Mike discussed the practical applications of a global namespace and file virtualization technology, and how we helped him save over $100,000 on hardware and labor in retiring and replacing an aging file server. Mike was also nominated for an award from the SNW coordinators (see below).
2. The launch of AutoVirt 3.0 took place on Tuesday of the show. This software is a substantial release that provides a platform for managing unstructured data. With new policies that enable tiering, archiving, disaster recovery and more, as well as interoperability with Microsoft DFS & DFS-R, this release offers tangible value to people who are centralizing storage, enabling remote backup, or migrating data to a NAS environment.
3. Managing unstructured data has become a critical task for many customers now because it is taking longer than 24 hours to conduct a daily backup. With storage needs growing much faster than network pipes, lots of companies are realizing they can never meet their business continuity goals without taking this data under a higher degree of management. And many have no idea where to begin.

Congrats to Mike Williams of Five Point Capital and to State Street

Mike Williams from Five Point Capital was named a finalist for Storage Networking World's (SNW) Best Practices in Storage Awards, under the category of ‘Technology Promise and Innovation.’ This was for his deployment of the AutoVirt solution described above. Although Mike didn’t win, the company that did win in his category – State Street – was actually doing something very similar. State Street had also deployed a global namespace and used it to implement a massive rearchitecture of their storage environment. State Street’s implementation took about 18 months and although a similar project run with AutoVirt would take on the order of weeks, we congratulate both Mike and State Street for their achievements and recognition.

I continue to stand in awe of Ray Kurzweil

On a personal note, my first job out of college was at one of Ray Kurzweil’s companies. Kurzweil Applied Intelligence, now part of Nuance after a string of acquisitions, developed advanced speech recognition software for navigating computer desktops and dictating documents. Ray gave a keynote address at the show where he talked about his life’s research – the meaning of information technology, the future of computing, and overwhelming empirical data supporting his conclusions. But the reason I stand in awe of Ray is because he is a true renaissance man: entrepreneur, author, scientist, and optimist. What can I say – I just admire the guy!

Windows data is business-critical data

The commenter wrote, “And quite frankly, Windows-based files are viewed non mission-critical data.”

Huh?  Windows-based files are becoming more and more critical as data grows and compliance requirements increase.  The importance of Windows data has little to do with the failure of Brocade, Rainfinity, and Attune.

In the healthcare industry, MRIs, XRays and microscopes produce large, mission-critical images that have to get stored.  That data needs to be stored at very high resolution, with lossless compression, virtually forever.  And as the disks that house these images fill up with new images, the old ones need to be moved around, processed, and archived. We talk to customers all the time who manually move these images around – it’s a huge time sink and one that has tremendous room for operational improvement by using AutoVirt software.

If you manage a small or mid-sized manufacturer, there are tremendous amounts of data to be managed from specifications to CAD to quality assurance and testing. Interrupting the flow of the assembly line has a direct impact on the bottom line, and data is a critical part of that process. To keep your line running, you’ve got to automate the movement of data across systems and functions.

Many of the law firms we are talking to have multiple offices spread out geographically, and with cases that can drag on for years, document management is a critically important task. But over the years, as case documentation has expanded beyond notes to include online photographs, audio and video, the pressures for managing that data are beyond the scope of what an individual office can achieve. So in order for a limited central IT staff to be effective across these remote sites, they need new approaches to data management.

What’s the point of file virtualization, anyway? 

File virtualization is a cool technology, but it is only a means to an end.  For AutoVirt, that end is data management.  Brocade, Rainfinity, and Attune were file virtualization products, and we’ve been told that they were bulky, expensive, and difficult to install and use.  Once installed, they changed the storage infrastructure – but they were often too difficult to use routinely for storage management tasks. 

From the beginning, AutoVirt’s developers have prioritized ease-of-use and configurable automation.  We think of ourselves of a storage management product – not as a file virtualization product.

Where share management ends and data management begins

The commenter on Crump’s blog declared that “Windows files (shares, etc.) are a pain but [a] solution that just manages them is a ‘nice to have’ functionality that can be done by other applications or manually.”  

Share management – creating new shares, setting up permissions, and moving shares between filers – may be an approachable task.  But these shares need to be part of the overall storage strategy, and it is here that managing them individually (or finding a collection of different tools to manage them collectively) becomes inconvenient and risky.   Data is growing fast, and so are business requirements around disaster prevention, data protection, high availability, and data archiving.  For some businesses, replication and archiving tools are still nice-to-haves – but such businesses are growing increasingly rare.  AutoVirt simplifies share management, and also provides automated replication and archiving. 

Data migration, defined as the relocation of shares from one filer to another, is easier with a global namespace.  Users and client computers use the namespace to locate their files and are therefore insulated from the actual physical location. 

Let’s go back to the problem of file server sprawl.  Suppose you want to move the contents of several Windows servers onto a new Windows server with better performance and higher capacity. 

A policy engine should handle data migration, but should also handle replication, failover, failback, archiving, tiering, retention, provisioning of filers and disks, and retirement of aging systems. 

Another checklist, this one for adding automation to your data management infrastructure:

• Is the policy engine integrated with a global namespace?
• Are the workflows you complete manually today available as policies?
• Can you automate job management and exception handling? 
• Does the policy engine scale easily as the amount of work varies over time?

UNC names are stored all over the place:

• in client computer shortcuts
• in network places, folders, and drives
• in embedded links in documents and databases

When a share is moved, finding and modifying every last UNC name instance is impossible or at least impractical.  A global namespace should address this issue by supporting all existing UNC names for as long as they are in use – possibly forever.  An old document’s UNC name may reference a filer that is long gone, but the global namespace must recognize the UNC name and know the current location of the data. 

It is possible to support pre-namespace UNC names if the global namespace is built using automated file server inventory.  This also presents an opportunity to use copy and paste to create additional namespaces, should you wish to create various organizational schemes for accessing the same set of documents.

Deployment of a global namespace should not be a “big bang.”  You want to ensure the parallel operation of old and new, making sure that the transition to the new global namespace is convenient and fast.  This is a very reasonable expectation: the global namespace leaves the underlying file systems intact, regardless of how client computers connect. 

This is a big subject so I will stop here and leave you with a checklist.  Assess any global namespace solution by asking the following questions:

• Does the global namespace support pre-existing UNC names?
• Is the namespace built by automated file server inventory?
• Is parallel operation – UNC name to DNS to filer, alongside global namespace referral – possible and easy?
• Can you create additional namespaces using copy and paste?
• Is the global namespace removable?
• Is the global namespace integrated with a policy engine?
• Does the global namespace get out of the path, once it has made its referral?
• Does the global namespace manage multiple copies?  Does it recognize which copy is the master, and can it designate a new master should the original master fail? 

I’m always thinking about new ways to express the conceptual stuff that AutoVirt does, and this formulation from Boles was helpful: file virtualization is about separating how data is accessed from how data is manipulated.

A couple of points from the interview that I’d like to reiterate:

-  File virtualization eliminates the single, direct access path from client to file.  This allows IT the freedom to manipulate data in the storage environment, without impacting client access.

-  Your file virtualization solution should be compatible with your storage management activities: it should allow you to migrate, replicate, tier, and handle failover.

AutoVirt’s philosophy for developing storage solutions is to solve problems through automation.  For us, file virtualization is the right way to enable powerful, intelligent engines for data management operations.  Through the implementation of file virtualization, IT administrators can manage the full data life-cycle from a single console. 

File virtualization is a cool technology, but technology for the sake of technology is relatively meaningless.  The truly exciting part about file virtualization is that it can be leveraged to remove bottlenecks and obstacles in data management. 

To explain further, let’s examine the better-known paradigm of web browsing.  Your browser presents all the web sites in the world to you.  The Domain Name Servers (DNS) of the world facilitates a direct connection from your browser to the corresponding website and it is thereafter out of the path.  Similarly, a global namespace in your company’s network will present all the file server shares in the company network to you.  The global namespace will then facilitate a direct connection from your system to some filer and it is thereafter out of the path.  Both are scalable for the same reason: their ability to get out of the way after their job is done.

A global namespace has four separate benefits, all of major importance. 

1. A global namespace helps users find data independently of file server sprawl by replacing multiple network places with a single, logically designed folder structure.  Questions like, “which file server has the 2005 datasheets for product X?” are made obsolete by a global namespace.  A global namespace may thereby increase productivity, lower support cost, and reduce the common tendency for individuals to hoard files.  (A global namespace should also give IT the option to maintain the existing user experience – that is, to leave existing UNC names unchanged.)
2.  A global namespace presents files independently of physical location in the network, thereby permitting continuous optimization of data and storage for business reasons.  With a global namespace, data migration is online and unnoticeable to users and client computers, making it easy for IT to pursue cost savings and service improvements.  Storage tiering, consolidation, pooling, and upgrades are trivial behind the veil of a global namespace.  Project costs are slashed and costly delays are avoided.
3. A global namespace has multiple replicated targets and failover/failback support.  Constant data availability creates a competitive edge for companies and is an important way to prevent loss of productivity.  Replication is cheap as capacity costs have plummeted, and multiple online copies may help solve backup problems.  Replicated data, preferably using multiple locations, provides a critical component of any disaster recovery strategy. 
4. Finally, a global namespace will provide access to the “closest” copy, saving network bandwidth, and reducing latency.

In conclusion, a global namespace provides all the management benefits of storage consolidation at a fraction of the capital costs.  The transition from individual file servers to a global namespace has been intrusive, difficult, and costly until AutoVirt automated the process.

Replace physical file servers with virtual machines using VMware or HyperV.   To the client, the file servers appear unchanged in number and content, so the transition is relatively easy.  Yet despite the reduction in physical devices – and concurrent savings in hardware – each virtual OS must be maintained as an individual server.  Multiple servers on a single piece of hardware means additional management complexity.  Furthermore, performance is negatively impacted by the extra layer of operating software, and more Windows licenses than necessary are used.  

2.  Windows file server consolidation

Windows file servers are generally utilized at less than half capacity.  Consolidating these underutilized devices means tremendous savings in management costs – if only the cost of executing a consolidation weren’t so high.  With shares growing at different rates and devices running at different capacities, deciding how to consolidate is a puzzle.  IT also has to consider the process of moving data – the more shares are moved, the greater the disruption to clients.

3.  Storage consolidation

Big NAS devices are cheaper to manage than hordes of small fileservers, but the dirty secret is that the cost per terabyte of mainframe NAS is much greater than the cost of commodity Windows hardware.  Furthermore, the raw performance of NAS can be limited by low-performing CIFS implementation.  The transition to NAS is costly because new and possibly better namespaces are created and users must be directed to new file locations. 

SANs do not relieve file server sprawl directly, but do make it easier to move and add volumes.

4.  Virtualize and automate

Combine a global namespace with policy-based automation.  The global namespace is a logical consolidation of storage resources: users browse a single file system for all the files they need.  The global namespace also provides IT the opportunity to manage resources and use storage tiering strategies, independently of and unnoticed by users.  Combine the global namespace with an automation platform, and IT now has a promising option for consolidation:

• Automation may reduce the cost and impact of converting to a global namespace: particularly if it continues to support the names used by user shortcuts and embedded links.
• Automation reduces the cost and effort of managing many inexpensive storage units to that of managing a single, consolidated NAS device.  With a virtualized file environment, it’s easy to manage many Windows servers.  It’s also easy to consolidate.
• Plain old Windows hardware – if managed by a software-based file server virtualization product – is cheap to purchase, upgrade, maintain and manage.  It is cheaper than either file server sprawl or mainframe NAS.

With a global namespace and policy-based automation, file server total cost of ownership decreases dramatically.

When you compare the TCO of your existing environment to the TCO of a refreshed or consolidated file server environment, consider these factors:

New devices will quickly pay for themselves, but how will you complete the transition?

It is time that the storage industry changes its focus from bits and bytes to service and usability.  It is time to make software do all the boring work, freeing the people responsible for making it all work to innovate.  My passion is automation.

While AutoVirt’s focus is on automating Windows storage management, my blog will frequently address the storage industry as a whole.