So now we have moved from the three-tier model to HCI. We have some thought leaders in that space one of them being Nutanix, another being VxRail, and then a slew of others that would be purchased by much larger entities over time. But let’s just focus on two of them and why they mattered and how they revolutionized the data center.
So the first piece of the puzzle when it came to three-tier management was how can your infrastructure be streamlined to make it more cloud-like in terms of consumption and at the same time remove some of the time overhead from your IT staff to make them more effective and happier overall? A lot of that comes down to just taking the complexity out of the way we were doing regular tasks and the worst part of that was updating and maintaining all those moving parts. With HCI, you then just had to worry about a chassis with nodes and drives all in one. Single point of support etc.
This sounds great and I made a sales career off of this but as a three-tier maintainer and an ESXi administrator, it was not what I wanted to hear. I felt my job was on the line because I was conditioned to think if I wasn’t working nights and weekends then I wasn’t doing my job and earning my paycheck. I rallied my team against Nutanix and against HCI in general. I refused to listen to anything my Nutanix SE said, so they decided to bring the team in for an EBC at the Durham office.
(Side note: I took this picture in the old Nutanix office in Durham, NC, and shared it via social media. The caption was “Time to hear about a bunch of bull s**t in the bull city.”)
Our Nutanix team brought in everyone and anyone that helped build Nutanix and the underlying architecture that was going to make my life better. We heard about everything and learned anything we wanted about Nutanix and how it worked. I sat at the head of the table with my arms crossed. Obviously not giving a damn during the 8-hour EBC about Nutanix or HCI let alone AHV which would replace my precious ESXi and therefore making me obsolete. I was a lost cause and was never going to allow this into MY data center. Luckily the choice was not mine to make.
Nutanix boxes were brought into my data center and installed within hours. Hours… not days or weeks like we were used to. ESXi was installed on them because it gave me a sense of worth and security. It was up and running and working. We had some RAM failures and a node failure and everything on it just kept running. This was foreign to me. We bought some more and installed them in one of our remote offices in Austin Texas. ESXi again and it worked just like we expected and it was awesome. The administrator there didn’t have to worry about it and if something broke we had the parts the next day and someone to replace them with no issues. Then my SE decided to truly challenge me and make me as uncomfortable as possible. To this day I cannot thank Tom Powell for doing this enough. He somehow convinced me to convert that Austin office from ESXi to AHV. “It’s one-click” he said, “and then you can convert it back to ESXi if you hate it.”
Now I want to take a second and set the scene: Raleigh NC sometime in summer where the temps outside reach 100 degrees Fahrenheit, and then add in the humidity and it’s awful. On Fridays, my former company would shut off the AC at 9pm (we never knew this as most of our maintenance windows were over at our hosted data center.)
So I scheduled a maintenance window and did it. It failed, the conversion stopped mid-node and just wasn’t doing anything. I opened a P1 ticket at 8pm and started to freak out, it had been two hours since the node froze and this was supposed to be 1 hour or less per node. My site was down what was I going to do, Nutanix was a failure so I had thought in my favor right? Wrong.
My team lead came back to the office to try and work through it while we waited for Nutanix support. We had a response but then the SRE went silent. 9pm hits and the AC turns off and the IT department got really warm, really fast. Our brains were racing, we self-escalated the ticket. Read that again… WE self-escalated the ticket. The next thing I know another SRE is calling me, we are on a zoom session minutes later my brand new account manager is on the call. Never once spoke to them as he was brand new to us as of that week. He was introducing himself to calm us down and let the SRE work. Absolute genius. Finally, around 11:30pm things were moving again, it was a simple restart of one of the backend services and the conversion happened flawlessly. By 1am all of the nodes were converted and all of the VMs were working and running as expected. Funny side note, our new account manager Jason Heinrich, had fallen asleep on the call. Nutanix and their amazing support had saved the day… well in my book they did.
I want to get to the setup for part 3 so to condense the rest of this. I converted everything to AHV. VMWare had sent us a count-down clock for the end of our ELA, we never told them we were moving to AHV. We let it count down and it just became a clock. (I have other stories around this clock for another time or over a drink.) I have started to evangelize AHV and Nutanix within every org of my company. I have DevOps willing to commit millions of dollars to Nutanix, from their budget, not mine. We go to .Next in New Orleans in 2018. Have the time of our lives. I talk to other companies about using Move to convert from ESXi three-tier to Nutanix AHV. I get to meet one of the people behind Move and thank him profusely. My sales team is trying to recruit me to Nutanix, but I hate sales teams and refuse to be on one. We get back home to North Carolina and shortly after we have an IT all-hands.
During the all-hands our director states “We are moving towards Azure and automation, if you do not start working towards this you will be out of a job within a year or two.” Well F**K!
So to sum this up, what does all this mean, it’s just my story of my journey from three-tier to HCI. How I got my life and time back. I have more stories about those items I will share. But IF I had stayed where I was and IF I had the tools that Nutanix has today what would that look like? Well, that will be part 3 The Future.
One response to “Infrastructure: the Past, Present, and Future (part 2)”
[…] had that same question back in late 2017 and early 2018. As I mentioned in Infrastructure: the Past, Present, and Future (part 2), I was in the process of switching from ESXi to AHV, and the company I worked for, had over the […]