Today is a new kind of birthday for me—it’s the day I started at Hyperic 6 years ago. It’s a journey that has been nothing short of surprising. My first meeting with then CEO Javier Soltero started the pace of what was to come: I had no idea it was an interview walking in, and was even more surprised when I walked out of that same interview accepting a job that same night. That sets the tone for the next 6 years.
It’s been an exciting journey riding the waves of technology changes these past 6 years, and sitting, in my opinion, in the perfect spot to see some of the biggest disruptors unfold. I have literally had the privilege to work with the most interesting companies in technology along the way. Most of the biggest companies on the web, many of the biggest companies on the stock market, and some incredibly innovative startups have used our software, partnered with us, and in two cases—acquired us. From this vantage point, I have seen some impressive changes in our industry along the way. Here are some of the highlights:
Virtualization would make systems management cool again. I’d been in enterprise software for ten years, including being a system’s integrator, and sysadmin work was a part of the job. For me, it was the worst part. Sitting in a freezing server room running scripts, reading logs, and tweaking configurations was my own personal nightmare. It was a 30+ year old discipline that was repeating itself with every new tech turn. So why would I take a job where that was all we would talk about? Because, during that first fateful meeting, Javier ran me over to meet the ever-inventive CTO Doug MacEachern. Hyperic had already focused itself squarely on web infrastructure, but Doug had created a plugin that would allow Hyperic to be able to see inside the host and guest operating system at the same time. The idea intrigued me—even VMware, the main virtualization provider I was familiar with at the time, hadn’t invested here. Their management capabilities focused on how the hypervisor itself ran. This was a gap.
It was also an interesting time for virtualization. At the time, the rationale for virtualization was mostly server consolidation. But a lot of companies were failing at virtualization, and massively. Some of it was due to inexperience with virtualization principles (people were virtualizing the wrong kinds of apps), but most of it was because systems monitoring and management tools had a huge blind spot. This is how I thought of it:
One of the first things I did at Hyperic was beg Doug to finish that plugin and release it under our open source product. We did, 42 days later on December 11th. This lead to the next big change for the industry.
Virtualization needed manageability for apps to be an explosive disruptor. As I said, at the time, big companies were still squarely focused on server consolidation and probably getting a C- at doing that alone. Put mission critical apps on virtualized servers? No way. They couldn’t trust it. But smart startups could. Amazon had started it’s cloud service in 2006 also and Google App Engine would soon follow. With just a credit card, brainiacs could start a software company in their mom’s basement and create software services that were relatively low cost and paid for themselves immediately. This pay-as-you-go model created a green field for innovators like Animoto, Netflix, and that BuddyPoke app that you used to “poke” your friends on Facebook. How could these guys build their whole business in the cloud? Well, to start, they built their apps specifically for the cloud. They didn’t try to move a legacy architecture there. They built their apps to be loosely coupled and horizontally scalable. They also started to look at monitoring and management differently. They didn’t ever care how the host servers were running, they looked at how their app was performing and monitored and managed it accordingly. So the little guys ended up using web infrastructure management tools like Hyperic to prove to the big guys that it could be done. This unleashed an exodus of apps from the datacenter to the cloud, which was the next big change that affected the CIO's books.
Shadow IT was going to become every CIO’s nightmare. I cited in an earlier blog post this fall that the new economics of the cloud operating model was showing CIOs that apps were landing on the cloud faster than ever before. And up to 30% of them were outside the purview of IT. Meanwhile, it has been proven that apps are 12% cheaper to run on internal clouds. No longer was it just easy for entrepreneurs to start their apps on the cloud, but line of business managers with a corporate card could just as easily start new business initiatives there. Most of the time, it was easier than going through IT to requisition hardware. With clear financial incentives, IT started to compete with public cloud vendors and build their own private clouds. But still, apps were leaking out to the cloud.
Virtualization software needed management as a service to compete. Predictably, since they’ve been in the virtualization game the longest, VMware recognized this years ago. SpringSource, who was leading the charge from the application development side on how modern java apps could be built and scale in the cloud, got to Hyperic first and acquired us in early 2009. This paved the way for VMware to acquire SpringSource and Hyperic just 3 months later, with the vision of making it easier to build, run and manage apps on cloud infrastructure—with the obvious goal of VMware powered clouds. Hyperic landed itself in a brand new organization known internally as the Management Business Unit (or MBU). The MBU has responsibility for the familiar VMware management tools like vCenter Operations, but also has an entire effort all around rethinking how not just the virtualization software is managed, but how apps are managed. Several acquisitions aside from SpringSource/Hyperic were part of this story including Integrien, Digital Fuel, and the latest DynamicOps. These acquisitions gave VMware a headstart and some solid talent to fuse together their vision of building out application management as a service for end users of the vSphere powered private clouds. The result? The new Application Performance Management suite that includes Hyperic and Application Director. While they do have some of the IP from these acquisitions and were definitely built on the backs of the talent acquired, these are fundamentally new software solutions that are pushing the envelope once again for virtualization. Out for just about a year in their current forms, these two products are fundamentally changing how IT operates. IT is now empowered to become a service provider, making it just as easy for internal users to procure computing resources and at the same time, making their service more attractive with built in management and automated, repeatable provisioning of their apps. These services, built into the vCloud infrastructure, are the capabilities that are finally steering Shadow IT back into the daylight of IT’s internal control and at the same time accelerating development even further.
So, 6 years later, I am excited and proud that I am still part of this story. I am also really happy that the 5 founders, Javier Soltero, Doug MacEachern, Charles Lee, Ryan Morgan and John Sachs as well as Jon Travis, one of the early engineers who really pushed Hyperic technology forward, are still very much part of this story. But I guess after 6 years of knowing that virtualization would make the topic of management cool again, I should also know that being part of the mothership of the virtualization movement with VMware would keep it interesting enough to keep us around for what’s next.