"Fabric Based Computing" the future of Fabric Apps?

If you attended the Gartner Data Center conference a week ago, you may have already heard the buzz on “Fabric-based computing”. It’s a fancy new label for Cisco’s entry into the market with the UCS platform. It’s almost a marketspace of one, but with the addition of the HP BladeCenter matrix and a couple other smaller players, 3 leaf networks and Liquid Computing. And that surprisingly, 30% of Global 2000 companies will run some form of it by the end of 2012.

Is this really a market? Or just a server bundling exercise being wrapped in revisionist PR. When Gartner first published about Computing Fabrics they were speaking to the emerging enterprise compute-clustering space, now populated with just a few remaining players such as Egenera, Scalent and Surgient. The others have died or been bought – some just for their IP – such as Virtual Iron/Oracle, Cassatt/CA, Platespin/Novell, Qlusters. The problem is that the Cisco modular blade server / network switch chassis is a world away from an offering that does virtualized server compute, shared system memory, automated/dynamic resource management, etc.

Also, I’d love to think that “Fabric-based computing” will enable a rebirth of the network-based application space. Again, products like UCS are just a server chassis (Cisco now offers both blades and rack-mounted servers) that run Intel processors and standard operating systems. This is decidedly different from how Invista or RecoverPoint runs actually within the network fabric, able to work with heterogeneous host operating systems. Not saying it’s better or worse, but just that it’s different.

Anyway, in Gartner’s study on the topic (just published, not yet available online), they point out that “Adoption will likely follow a similar pattern to that of the early years of the blade server market.” This is because UCS is primarily a blade server. No need for a fancy new name. And no need to invoke the spirits of Fabric Application or Compute clustering.

New Video on Fabric-based RecoverPoint

A product update video about RecoverPoint on Brocade was posted today on youtube featuring Rick Walsworth of EMC.  It provides an update about this leading fabric-based Replication and Data protection solution.  Known as the product of choice for enterprises and large organizations protecting heterogeneous storage envuronments, the offering now works with Virtualized environments and supports Brocade's newest SAN directors, the DCX and DCX-4S. Highlighted in this video are how RecoverPoint works with Brocade fabrics, protecting Virtualized environments, and EMC / Brocade technology synergy.  The video is available here, and more info on the solution is available at www.RecoverpointOnBrocade.com.

Fabric-based Replication In the News

A recent Storage Magazine story addresses network or Fabric-based storage management including Replication. The article, "The pros and cons of network-based data replication" by Jacob Gsoedl specifies, "Network-based replication combines the benefits of array-based and host-based replication. By offloading replication from servers and arrays, it can work across a large number of server platforms and storage arrays, making it ideal for highly heterogeneous environments."

The writer clearly gets the core benefit of fabric apps for enabling the management of mixed storage environments. It cites a number of replication and virtualization products including EMC RecoverPoint, IBM SVC and HP SVSP.

To learn more of the specific features and benefits of a leading fabric-based replication offering, go to www.RecoverpointonBrocade.com.

The Golden Age of Fabric Apps

I found myself describing Fabric Applications as being in their "Golden Years" the other day. For being completely off-the-cuff, it's not a bad description. They are very mature and in that sense, people understand what they are and the value they bring. For instance, Network-based storage virtualization is still written about as an efficient approach to deliver the benefits of virtualization across a cross-vendor or heterogeneous storage environment.

Having said that, we're seeing more developments in array-based storage such as FCoE. And we can continue to expect more virtualization and replication capabilities consolidating onto the array, esp. with new modular designs like V-Max. Network-based offerings from IBM SVC and HP SVSP seem to defy this trend but those roadmaps should converge towards the DS8000 and EVP respectively at some point, as HDS offered the market solid validation of array-based heterogenous virtualization with the USP-V.

Don't get me wrong - I'm not sounding the death knell of network-based storage services. I'm guessing just the opposite: more intelligence being built into converged networks leveraging advanced standards such as FCoE, CEE and TRILL. And also continuing to provide common data management services and offload that allow applications using the fabric to be that much more efficient.

So for now, in addition to fabric apps, I'll be spending more time in the cloud space. Please follow the conversation on my new CloudItch blog (www.clouditch.com).

Congrats on V-Max

EMC had big news today thanks to their virtual launch of the new Symmetrix V-Max. It has garnered broad industry coverage, as well as its share of competitive brickbats.

It's a revolutionary break from the old-line monolithic array. Based on x86 components, designed with modular 'engines' to enable more granular scale-out, and built-in automated tiering makes it more like an offering from a nimble innovator like Compellent or 3Par. But with scale-up to a claimed 3PB and "tens of millions of IOPS" this is still gold-tier enterprise storage.

EMC's Chuck Hollis paints a typically empassioned view of the technology. It includes storage virtualization features such as pooling data across multiple arrays, non-disruptive data mobility and I/O load-balancing. He listed architectural approaches customers have to achieve Storage Virtualization: "today, we've got three different approaches to putting storage virtualization in the network: (1) use a server appliance (e.g. IBM SVC), (2) use an array controller (e.g. HDS USP), or (3) use an intelligent switch (e.g. EMC Invista)." Forgiving the over site of "host-based" (e.g. DataCore), this list does cover the typical enterprise, large service provider or government agency considered set. There are a number of choices each with its pros and cons.

And despite all the greatness of V-Max, the Storage Virtualization buyer often needs to address issues of heterogeneous storage resources, performance and vendor lock-in. Where these are issues, a fabric-based approach is still the best way to go. (And if you're an EMC shop, you can still make your sales rep happy by choosing Invista.)

Congrats to the Symmetrix team on the launch -- we're all eager to see how the new V-Max vision rolls out over the coming months.

Virtual Storage as well as Coffee Cups

The report from SNW: still a big show but not quite as large as last year. Cost cutting was top of mind, with one attendee sharing that his company eliminated styrofoam coffee cups to save $70k!

Storage news included the apparent maturing of Storage Virtualization, with 76% of attendee orgs either having already deployed or expecting to deploy a solution by next year. Companies are realizing savings from Storage Virtualization through efficient resource pooling, reduced management complexity and the ability to manage more data with the same staff. One company shared that they had 15PB all on a performance SAN, and wanted to reduce cost through tiering to cheaper arrays -- a compelling use for Storage Virtualization.

With the maturing of the category, the key players will stabilize their respective positions. But there's a potential disruption on the way, because the market still hasn't found an equilibrium in terms of form-factor, with shares split between array-based (e.g. HDS USP), appliance-based (e.g. IBM SVC and NetApp V-Series), Host-based (e.g. Datacore) and Fabric-based (e.g. EMC Invista). Expect to see innovation in how network-based technology works to improve the performance of these and other storage apps...

Pay-as-you-go Computing

It's exciting to see as both vendors and customers become more evolved with this IT business, that the nature of our transactions change. The recent buzz about the expected explosive growth in Cloud computing has not-surprisingly coincided with continued IT cost reduction pressures. Some is certainly due to the current global economic crisis, but the trend has been in play for a long time. What has gone from an exercise in procuring something cool and breakthrough, has become buying something basic and vital, like milk. I've taken a stab at capturing some key aspects of this evolution to cloud computing.

What I'm seeing as especially significant is how we as infrastructure technology providers are going to have to package our solutions in a more pay-as-you-go fashion. Very much like how we all buy our power at home - it sounds like a no-brainer, but let's say you make your living selling network hardware. How does that work? And as the customer, what do you expect this utilization-based arrangement to look like? Or maybe the real takeaway is that you don't want the arrangement at all. You want email, a way to track your sales pipeline, and a means to do book-keeping, and don't want to have to staff experts to determine if the best way to do this requires Fibre Channel, Ethernet, FCoE, CEE, or M-O-U-S-E.