Have you ever heard the expression “The only thing not being created anymore is time”? I used to hear that often growing up. Back then it made little sense to me but as I got older, married, started a career and then had kids it seems that the hours in the day are shrinking and shrinking. It’s amazing !! This has drifted over to my career. It seems I have more “work” projects then I have hours in the day to get done, which means I spend lots of after work hours trying to play catch up. When I worked in the datacenter, this was always the case as well. I learned a new phrase a few years back called Non-Recoverable Engineering time (NRE). NRE can be summarized with “All that crap that you need to do that has little to no value to the bottom line”. I think we can all agree that our lives (personal and professional) are full of NRE’s 🙂
Remember back in the good old days, when we seemed to have all the time in the world? If you were “nerdy” you’ve probably at some point in your life built your own PC. Remember, all the fun it was to do the research on the best mobos, procs, memory, disk drives etc. You probably spent weeks debating the latest processor and the cost difference between model numbers. You would pull together the list of things you wanted, then figure out your budget doesn’t even cover half so you start trying to figure out how to swizzle all sorts of things around. Then you would spend a marathon weekend pulling it all together hoping everything worked, looking for updated drivers, cursing up a storm and then after a week (or two) everything was running great. It was probably a month’s worth of effort !! I did a few of these in my life so I got pretty good at it. In fact, today I could knock it out in about a week’s worth of time (research, order, put together etc). But I haven’t built a PC in 10+ years, in the grand scheme of things on my “ToDo” list I just couldn’t justify the time and effort it took to pull all of that together. Building a PC was what I would call a MAJOR Non-Recoverable Engineering (or Daddy) time suck.
(Stole this picture from Kenny’s blog post 🙂 )
So fast-forward to today. I really needed (wanted) a new laptop. My work one just wasn’t cutting it and I’ve wanted to jump on the Apple bandwagon for a while. In this case, I decided to go to a MacBook Air. I went with the 4GB memory, 256GB SSD HD and upgraded processor because those were my choices. I could have chosen a bigger/faster laptop but this was more than enough to meet my needs and it’s perfect for all the travel I do. This system had a VERY low NRE hit to my life!! In fact I was out with the family when I bought it. My 9 year old son and I went to the Apple Store while my wife and 7 year old twin girls went to Justice or some other “gurlie” place like that :). Anyway, we walked out of Apple Store and sat down in the mall and started playing around with the laptop. In less than 5mins I was connected to my Mifi and was downloading tweetdeck and all the other software tools I wanted!! It was awesome. I took what would have been an all week affair, locked in my home office putting everything together, to less than an hour to go to the mall, go to the apple store, pay for it and start using it.
So hopefully you are seeing my point. It really boils down to how to reduce all of that “Non-Recoverable Engineering time”. For those that are struggling with “What exactly is a Vblock” or “Vblock vs NetApp FlexPod”or “what is the value of Vblock” then consider my MacBook Air a sort of Vblock. I can choose from a few different models (Air vs. Pro) but at the end of the day it’s about getting more of my “time” back. Sure I have the core competency to build my own computer, just like your teams can build their own reference-architecture solutions (VMware + Cisco UCS + EMC or NetApp) but in the grand scheme of things, there are probably a ton of other projects they could be focused on. Not to mention, one of the LARGEST NREs is the care and feeding of the system once it’s installed. Firmware upgrades are an example I discussed in this blog. The time and effort it takes to do the research each time Cisco, VMware or any other device/hardware manufacture updates their firmware can be a HUGE NRE time suck !! With Vblock, VCE has done all that work for you. What you get is a complete tested, integrated, validated, 1 single pane of glass to manage, 1 support number converged infrastructure solution !!
A picture is worth a thousand words right?
Tried and true Reference-Architecture – This could be EMC + VMware +Cisco UCS and/or NetApp FlexPOD but either way, we’ve been doing this for many many years. Best of breed technology (Some assembly required) – this is what shows up in your datacenter.
Or you can consider how the Vblock arrives at your datacenter.
Any questions ? 🙂
vtexan.com 
I think you mean Non-Recurring Engineering Costs (http://en.wikipedia.org/wiki/Non-recurring_engineering).
Hey Jim – thanks for the comment!! I looked at that while I was pulling this blog together. I’ve heard it “Cost” and “time” – i should probably create a Wikipedia page for “time” 🙂 I think “NRE Cost” can be viewed over the life of the project, and I think “NRE Time” can be viewed at a more granular level – if that even makes sense. it does in my head if that matters 🙂
Hi Tommy,
I respect both Vblock and FlexPod. Your example about the shipment is logistical only. Vblock is shipped a configured product while FlexPod shipped in parts. But who cares who is it shipped when the VAR or Partner intergrates it for you (as a product)? Just my $.2
Yyaazz,
Thanks for the comment!!
So this is just my opinion –
If you look at the process as a whole, from researching the solutions parts needed, to procuring the parts, to installing all the parts, to the ongoing care and feeding (firmware upgrades etc) is a ton of non-recoverable engineering time. At the end of the day, you need to support an Application and you need it to have X uptime and Y performance.
Just like we used to build our own whitebox servers (partners/customer) for our datacenter, that transitioned to a couple of clicks on the website and ordering a server, I think we are moving towards that same approach for integrated converged infrastructure solutions.
Don’t get me wrong, I love racking and stacking as much as the next guy, just like I love to build my own PC’s but at the end of the day, on the list of things that need to be accomplished (partner or customer) I feel like I can have a better use of my time focused on the things that have a large impact on my life (personal and professional).
Tommy,
I can agree some points and some points not 🙂 I’m working for a partner certified for FlexPod and soon for Vblock (VCE channel), so I can evangelize both as I don’t work for EMC nor NetApp.
Somebody has to take time. From the customers point of view, who ordered a turnkey solution he doesn’t care, it’s not his time. Even it’s a Vblock or a FlexPod somebody has to get the parts, upgrade the firmwares, install the softwares, etc. It doesn’t matter who do this, partner or vendor, somebody has to do this.
cheers
One of the things I’m hearing from VAR partners and SIs is that their customers are attaching almost no value to the “rack and stack” portion of the implementation project. This means that the partner usually ends up eating that part of the cost, and either has their high-end architects doing menial labor or has additional people on the payroll just to handle this part of the project. Either way, partners want to head up the stack, not down. Having the infrastructure come ready to power on means they can spend more time engaging with the application owners and line-of-business owners and less time sitting in a datacenter running cable.
You make a good argument Tommy, but unfortunately One Size doesn’t always Fit All. Consider your MacBook Air analogy: What if you needed to attached an eSATA drive? Or USB 3.0? Or an ExpressCard? You can’t — the MacBook Air doesn’t give you those options. What if you wanted to upgrade your SSD drive to a larger capacity? Sorry, you can’t do that, it’s proprietary. (Can you say “vendor lock-in?”) Also consider the task at hand: What if you needed to do heavy duty PhotoShop work? PhotoShop CS5 uses 14GB of memory all for itself, so clearly the MBA with it’s maximum memory of 4GB wouldn’t cut it.
The point I’m making here is that there is always a situation where a cookie cutter solution doesn’t provide enough flexibility, and there will always be a place for best-of-breed solutions. This isn’t a dig at EMC, quite the contrary, just a reminder that EMC also sells best-of-breed solutions, even though vBlock seems to be getting all the press these days. Actually I think EMC is genius in this respect, as they can lead with cookie cutter and close with best-of-breed once they better understand the customer requirements; either way they win.
(Disclaimer, I work for Brocade (an EMC partner), but this post is my own and does not necessarily represent the views of my employer.)
Charles – thanks for the comment !!
So, I couldn’t agree with you more and I don’t think you will find anyone within VCE or the parent companies EMC, Cisco, VMware or Intel saying that Vblock is the only way. There are choices within the Vblock family for companies/and application workloads of different sizes. Not to mention, if the Vblock is not a good fit, then the tried and true Reference-Architectures is the answer !!
At the end of the day, customer do want/need choices. BUT just because something is rigid doesn’t mean it’s a bad thing. Just like there are many buyers that refuse to buy Apple products because of their rigorous proprietary nature, it’s not slowing apple down on creating some great products nor people that want to buy their products!! Look at the iPhone/iPad as a great example. I’m sure you know that the iTunes store passed the 10billion application download mark a few weeks ago and yet people complain constantly about how ridged their products and applications are. The iPhone/iPad moved from being a PDA or tablet to being a platform to run applications on.
At the end of the day, it’s about the applications, not the underlying hardware. Just like my choice of a Macbook Air had more to do with needing a solution to fit my travel needs, others would choose the more powerful MBP product line to do their Photoshop work. VCE has Vblock 0, 1 and 2 and it depends on your application workload !! Then again, you could go to NewEgg and order the biggest, baddess motherboard, processor, SSD drive etc etc to meet your needs and build it yourself (or pay someone to build it) !!
Also, just to be clear. Vblock is NOT a reference-architecture. It is a product. With a product comes all the things you would expect. Ease of deployment, ease of procurement, single pane of glass management, robust roadmap, single support number/contract, and a single company VCE that owns that WHOLE process.
Again, it’s not really about the hardware, it’s about the applications !!
Hi Tommy, sadly I have to say I dont agree with the assessment that you “don’t think you will find anyone within VCE or the parent companies EMC, Cisco, VMware or Intel saying that Vblock is the only way.” I can say I have been in many meetings where vBlock was the ONLY product being discussed as a solution for a customers needs. While we are all still debating the merits of vBlock (as evident in this posting), the parent companies along with VCE should consider vBlock as just one of the other products on the proverbial shelf. Unfortunately, I’ve been in enough meetings to see customers get so beat over the head with vBlock they dont really want to see any variation of the V(vmware), C(cisco) or E(emc) letters together OR apart!
On another note, I find that partners (at least the one I work for) dont necessarily just want to be the guy who puts the NewEgg parts together for you. We want to be the BestBuy (GeekSquad?) dude who knows all the laptops (Apple, Dell, HP) and, based on your individual specific needs, can point you to the laptop that works best for you. Maybe it IS a MacBook Air, great! We can sell that to you (with an extended warranty . . . am I right? 😉 But maybe you have a USB3.0 hard drive that you depend on and you cant afford to just go buy the Apple equivelant. Well maybe this Dell is a better suit. Or, just maybe, NewEgg is just the right fit for you . . .we can sell it and build it. I dont think partners feel threatened or afraid of vBlock and what VCE brings to the market; we just want the MacBook INSIDE the BestBuy . . . not a MacStore across the street!
By the way, great to meet you today. 🙂
Tommy. Great Post! I think it’s important to make this distinction about the differences between a product and reference architecture. I would like to add that, in my view, it’s very easy for us, as technical professionals to pick things apart and start tweaking. We have been doing this for decades now and it’s just natural to us, as we see things other people may not. However that doesn’t mean that by tweaking things we are adding value that actually justifies all the man hours and risks associated to the business by doing it. I think we need to step back and look what people that will actually use what we are creating are looking for. Coming from an application delivery background I say it’s predictable performance and that’s what a Vblock provides. The moment I give my application owner an Infrastructure that I can guarantee that his/her application will perform well and meet it’s SLA to the business it’s all that matters.
Tommy,
I’m an HR guy at EMC, and in an attempt to better understand the technology side of our business, I’ve started reading your blog (among others). This post really hammered home the value-proposition that VCE and vBlock bring to our customers. Loved the analogy with the MacBook Air (including the “this isn’t for everyone” discussion in the comments section).
Thanks for helping a non-technical dude better wrap his head around this stuff!
– Chris