The Virtual Workload Silo-ization Syndrome (The Webscale Imperative)

Many data center environments are plagued today by what I’ll call virtualized “silo-tary confinement.” This means that virtualized application workloads are often segregated or containerized within discrete pools of compute, storage, and network resources. For example, a virtual desktop environment may be configured with server, SAN/NAS storage, and networking hardware, while an Exchange infrastructure may utilize a totally separate pool of virtualized resources. This tendency to silo virtualized workloads drives up costs, increases management complexity, and hinders business agility; flying directly in the face of why organizations virtualized their server infrastructure in the first place. 
Topics: nutanix virtualization computing infrastructures Hyper-converged Hyper Scale

Top Five (ish) Storage Predictions for 2015

With the New Year upon us, it is time to make some predictions about 2015. If you have been paying attention recently, it will come as no surprise that the storage industry is in the midst of a dramatic transformation. With emerging technologies such as solid-state, software-defined storage, hyper-converged, and cloud storage all primed to increase adoption, 2015 is poised to be a fascinating year. Any of these new technologies could eventually disrupt the industry and change how companies do business. The coming innovations can almost help me forgive the fact that I won’t be riding around on a hoverboard (even though the next best thing looks like it may be in the works), let alone a flying car.

 With all this innovation underway, what will the storage industry look like in 2015? One of the things to remember when forecasting storage technology is that the market moves at a conservative pace. Protecting data is an important job and most organizations take that job very seriously, as such dramatic shifts can take a time to build up momentum. When they hit the tipping point, however, the industry can transition very quickly. All of the technologies mentioned above are building momentum, so which will reach the tipping point in 2015? Here are some of my predictions:

  1. Solid-state for everyone: I am not including this as a prediction, since saying solid-state will increase in adoption is about as risky as saying the sun will rise in the East. What I am going to predict, though, is that 2015 is the year we stop talking about solid-state as some sort of new emerging technology. It’s here, it’s fast, it getting more affordable, and nearly everyone is already using it.
  2. Vertical integration driving new storage players: With the rise of flash both in terms of adoption and affordability, flash suppliers are building more fabrication plants. And with all the investment, flash suppliers are starting to look like they may want a higher return. Additionally with the market demanding lower cost flash, vertical integration is one way to reduce prices. Whether through internal development or acquisition, such as SANdisk’s acquisition of Fusion IO last year, I expect that we will see more component providers entering the enterprise storage market.
  3. Showdown between hybrid cloud and cloud security concerns: Two ongoing trends will likely meet in a collision course this year. On one side the hybrid cloud--storage vendors have been talking about it for a while and have started offering multiple solutions designed to migrate data offsite, potentially lowering storage costs while maintaining the service levels of on-premises storage. One the other side, reports of data outages and security breaches with cloud solutions continue to dominate the media. While I expect a number of organizations to adopt a hybrid cloud architecture, I expect that the security concerns will be a major, if not the biggest, hurdle for those solutions to overcome and may even fuel demand for wholly on-premises content storage solutions in the near term.
  4. Software-defined storage will mean something: In 2015, software-defined storage will continue to dominate many industry headlines. And while I do not expect adoption to increase too significantly this year (maybe next year or 2017), I do expect we will all start to agree on what the term means. With multiple vendors leveraging the term for different offerings, confusion has been rampant. However, just as the greater familiarity with cloud technology allowed for a better consensus to emerge around the definition, I expect the industry to at least converge to some extent on a definition for software-defined.
  5. Server vendors will place more emphasis on hyper-converged: One area where software-defined storage could likely see an increase in adoption over the next 12 months could be in the form of hyper-converged solutions. And while a number of startups have led the change, I expect the server vendors, HP, Dell, and Lenovo, to start driving more of the adoption, whether the solutions are EVO:RAIL, reselling a partner's technology such as Nutanix, or something developed in-house.
  6. Start to investigate the software-defined data center hyper-scale model: This is the year hyper-scale (or the software-defined data center) solutions start becoming a discussion point. I don’t, however, expect the majority of organizations to start deploying or adopting these solutions any time soon. With many of the technology pieces already available for the non-Googles and non-Facebooks of the world to heavily leverage commodity hardware, I fully expect a vocal segment of the market to start asking, “why not me?” The complexity, however, is still on the high side for the typical enterprise, and that will hold back major adoption, but the pieces are there to start the investigation.
Topics: software-defined storage solid state storage Hyper-converged Hyper Scale