Home » Availability » Where Is All This Going? The Data Center Transformation Around VMware/Virtualization (Part 1 of 3)

Where Is All This Going? The Data Center Transformation Around VMware/Virtualization (Part 1 of 3)

 

  • Part 1: First Wave: Server, Storage and Availability
  • Part 2: Second Wave: Networking, Security and Management
  • Part 3: The End Game: ITaaS and the Private/Public/Hybrid Cloud

I had the opportunity to work at VMware from 2004 to 2010 as their first director of global channel marketing, programs and enablement.  It was a great experience and it gave me the opportunity to work with the alliances ecosystem and see how many of the technologies of the modern data center fit together, with virtualization as the gunpowder and the glue. This experience and my work with clients in my consulting practice over the past few years  provides me with some insights on how these changes fit together and some perspectives on the question,“Where is all this going?”

Let’s start with a look at the first wave of change that VMware and virtualization brought to the data center.  For each of these technology ecosystems I will discuss my observations on how these changes occurred, the impact on vendor ecosystem and the implications for customers and vendors in the future:

  • Server – The impact of virtualization really started with VMware’s first bare-metal hypervisor, ESX in 2003/2004.  Over time, VMware adoption has driven the need for multi-core chips, high performance servers, blade servers and ultimately virtualized architectures like Cisco UCS.  After all, what applications would use the capabilities that are available from Intel octo-core chips, if customer only ran one application on a single server? In the early days (and even sometimes today) there was always a question, “Why would IBM/HP/Dell support VMware if it reduced server count?”  The answer was that the product solved customer problems.  There were plenty of squabbles at particular accounts over this issue, but ultimately, customers won out, and got reduced cost, better manageability and a better platform for IT – a good lesson for vendors as customers adopt networking and storage consolidation technologies (that introduce some similar trade-offs for vendors…)
  • Storage – Has always been closely associated with Virtualization – reports in the mid-2000s from analysts like ESG found that most virtualization projects resulted in new shared-storage with a significant number of these sales to customers who were buying their first shared-storage. This connection really accelerated in 2006 when VMware began to support iSCSI and NAS storage and led to a parade of new vendors and storage architectures (like EqualLogic, who grew rapidly and was bought by Dell and LeftHand Networks, who grew rapidly and was bought by HP).  Companies like NetApp and EMC needed to defend their share and also saw a huge opportunity in this growing market, and they hitched their sales and marketing wagons to being seen by customers as “the #1 storage vendor for VMware environments”.  All the major storage vendors have been on a journey expand their architectures to address different tiers of storage, and to become more software-driven. This has led to rapid adoption of innovations such as de-duplication, and storage-tiering that customers needed to support the diverse needs of their (often virtual) environments.
  • Backup, Availability and DR– By 2004 it was already clear they customers needed backup and recovery options that were designed for virtualized environments, and backup vendors like Vizoncore (bought by Quest Software, then Dell), Veaam and others filled this need.  These new products focused on he needs of virtualization admins and caused share erosion from traditional heavyweights like Symantec, IBM and CA.  Traditional vendors took a product-centric view of the market and failed to recognize that other vendors were winning share by addressing the needs of the virtualization team (not the backup admin). Starting in 2006/2007, expanded storage compatibility, built in availability and simplicity of the VMware platform, and Site Recovery Manager product (SRM), made it viable for customers to think of VMware as a platform for providing high availability for all their applications (rather than developing a HA/clustering approach for each application that IT supported).  During this period, I heard a number of VMware partners say that if they had to choose only one benefit for VMware, it was the ability to enable DR (even more than the ability to reduce server count/costs!).  DR drives even more storage, so this trend pulled storage vendors and their substantial channel network closer to VMware, as they learned to position VMware, AND sell more storage.  Today all of the ecosystem vendors provide deep functionality for VMware environments, but will they stay ahead of the curve and incorporate functions required for SDDC and the next wave of changes in the future?  History shows that some will fall behind…
  • Converged Infrastructure and New Storage Architectures – Starting in about 2008/2009, we’ve seen a convergence of physical infrastructure across server, storage and networking into a single platform. This started with VCE/vBlock, and FlexPod from NetApp, but more recently has continued with hyper-converged infrastructure like Nutanix and Simplivity.  Today there are also new storage architectures from companies like Tintri and Pure Storage and the explosion of flash (from companies like Fusion-io, Violin Memory and XtremIO) in every part of the architecture, both as storage accelerator, and as primary storage for demanding use cases. Last fall, VMware announced a Beta for its new virtual SAN (vSAN) product and has positioned vSAN as eliminating the need for separate shared storage (in some use cases).  They also have begun to talk about their offerings as part of a larger industry trend toward  “Software-Defined Storage” (SDS).  For more information on the SDS vendor landscape, check out Torsten Volk’s excellent blog post. Throughout this transformation, the storage market has continued to grow faster than other technology segments and the base architecture is still in flux, ten years after this revolution began. This is an interesting lesson on how long these transformations take, and a reminder of how changes can impact slow-moving vendors.

What is in common for the transformation path for each of these ecosystems?  Dramatic change over a 10+ year period.  Change in the products, the overall architecture and how products were bought and sold.  Vendors that put focus on being ahead of this change flourished.  Vendors who tried to market their way out with existing products were left behind.  In next week’s post, I will talk about the second wave of change – networking, security and management, and the associated vendor ecosystems.

I welcome any comments below — And make sure you “Follow” our blog (look for the “Follow” link on the upper left) and have your say.


Leave a comment