By: Dan Newton, CEO – HiveIO Inc.
I have spent over ten years working for managed service providers, providing hybrid cloud services to some of the world’s largest organizations and best-known brands, and I have experienced firsthand some of the challenges that MSPs face today.
Our teams provided managed services to hundreds of thousands of customers. From small-medium businesses to mid-market, and enterprise. Rarely did two customers have the exact same requirements, infrastructure, or applications.
The vast majority of the solutions across data centers globally were based on three-tier architecture. This meant scaling out teams of specialists – network, virtualization, storage, Windows, Linux, database, and security teams to name but a few.
Management layers had to be added on top to deal with the ever-growing number of vendors. Throw in the mix the public cloud providers, for those hybrid customers, that were coming out with 1,700 new features and functions yearly and the layers of complexity were ever increasing. Not to mention the challenges this brought to procurement and pricing teams.
Most customers did not actually need this level of complexity. Yes, some customers were delivering custom developed applications that were revenue generating that they containerized, had automation, continuous integration and continuous deployment methodologies for – but these were, and remain, few and far between.
And if a customer wanted to deploy virtual desktops – I was the first person to say no. Too many vendors, too expensive, not enough margin, and too many handoffs between different teams.
The incumbent Hyperconverged Infrastructure (HCI) vendors promised to change the landscape, to simplify the solutions, remove the complexity – and were in my office pitching on a weekly basis. Some just bolted a new module onto their existing catalog, charged you extra, and added yet more confusion to their licensing model. Others had a good product and good support, but you had to meet their specific and complex hardware requirements to get performance which meant exorbitant costs. Between the complex licensing, and the significant CapEx requirements, I could not find a path to HCI across the data centers that allowed me to move from a team of specialists to a team of generalists – and reach my ultimate goal, reducing OpEx, reducing CapEx, and improving customer outcomes.
This is why Hive Fabric clicked instantly.
A solution that provided virtual server and shared storage capability, that had its’ own connection broker and served virtual desktops out of the box with a bare metal installation on any x86, and no requirement for other vendors. All done at a click of a button, no specialist teams required.
Not only was this packaged in an easy to consume predictable procurement model – by physical server by month – but it also answered one of the other hardest questions in the data center – data.
We always knew data was key. Primarily for monitoring. Spend a significant amount of your annual budget to get all the data in one place, normalize the data, and act on the data. If you were lucky you knew there was a problem at the same time your customer did if you were very lucky, you knew before they did, and if you were extremely lucky you informed them before they noticed there was an issue. The ultimate goal was to fix an issue and then inform the customer – no business impact and a very happy customer. Just getting to this scenario took significant investment, both money and time. Taking the next leap to data analytics and correlation, let alone AI and ML, was a whole other investment conversation and justification.
This is where the Message Bus in Hive Fabric comes into play. Our Message Bus reads, and provides access to, all actions and related data that occur within the Fabric from the foundation of the hardware, through the hypervisor and all the way up to the applications inside the VMs your running. All this data is in one place, normalized, instantly consumable and AI ready.
A Hyperconverged Fabric solution that provides VSI, VDI, and shared storage in a single bare metal deployment, that can be supported by a single team of generalists and is AI and ML ready – now that I would have taken.
If you would like to find out more about Hive Fabric, why not request a demo.