Aren’t we human beings curious things? We hold on to the past with one hand while grasping at the future with the other, completely oblivious to the present that is happening before us.
This post is my insight into how things are changing now in the present IT industry and how Protocol Passionistas are both risky to hire, own or loan for any organization; and if you see the characteristics of a Protocol Passionista in yourself, then it’s time for you to change.
In the IT world you meet professionals (small p) who have grasped hold of technologies and defend them like their (professional) life depended on it. You don’t have to look far for this in virtualization with VDI desktop protocols (ICA vs. RDP vs. PCoIP etc) or storage protocols (NFS vs. iSCSI vs. FC). Just walk around any data center with one of these professionals and ask them “Why did you choose <insert technology>” and it’s like you are asking why they chose their wife, like there’s some kind of inferred criticism, like questions and inquisitiveness are bad. Why is this? When the defensive attitude is related to protocols, I negatively refer to these professionals as Protocol Passionistas.
Don’t mistake a protocol vendor for a Protocol Passionista – they are a special case when selling that protocol. I’m talking about independent consultants, contractors and permanent staff.
If you want an example of the opposite of a Protocol Passionista, take a look at Scott Lowe:
- He’s an independent consultant (ok, he works for a reseller, but he’s not tied to one vendor or tech).
- He can have a balanced discussion with you about any virtualization technology.
- He can have a balanced discussion with you about any protocol (just read his book Mastering vSphere 4 to see this in evidence).
More than anything, Scott is a fully signed up member of the Inclusive Community: he rightly advocates the “And not Or Approach”, choosing the best-fit protocol (not based on any tech-religious beliefs, but scientific data) but also running more than one to suit multiple use cases. Scott constantly works out of his comfort zone and is these days familiar with Cisco’s data center virtualization strategy and products, like the Nexus family of converged switches (LAN and SAN in one) and the converged compute environment, UCS. In fact, Scott is a blueprint for the new multi-skilled Data Center Architect.
Where are you and the other folks you know on the Protocol Passionista scale?
- I believe in picking the right tool for the job based on latest tools and data.
- I am a firm believe in specific protocols for specific jobs based on past experience.
- I only ever use <pick protocol>.
- I follow the crowd.
The lower down that list you are, the more worried you (and your employer) should be. Don’t just answer this subjectively, look at the last project you did. Now answer the question again. Where are you really?
Look at storage protocols: Fibre Channel over Ethernet (FCoE) on Cisco’s Unified Computing System (UCS). People new to UCS immediately equate UCS = FCoE, and assume it’s “FCoE or NFS or iSCSI or FC”, when in fact it’s all of those – AND not OR. And it’s changing each quarter (getting better)!
UCS makes a nonsense of storage protocol passionistas because you can use anything and for many workloads – FOR THE MAJORITY OF USE CASES ANY PROTOCOL WILL DO. There, I said it. Stop over engineering and JFDI people! The most important goal of any project is to deliver something fit for purpose on time and within budget. If you can reuse existing stuff to reduce costs and do things faster, you are a hero to business. Over engineered, late and expensive is a career limiting move (although vendors might love you for it).
Look at remote desktop protocols: VMware just released View 4 with software PCoIP to compete with ICA, and surely that’s a great arms race for customers, driving down costs and increasing innovation: perfect! But I often meet technical folks with arms crossed, defending ICA as the most perfect protocol that “why would one consider anything else?”
That kind of closed mind is like nails down a chalk board for me, and I’m sure it is for the organization leaders who hire people like that. If you’ve done recent, relevant and scientific comparison with an open mind – ie. actually done an evaluation, and not one based on a vendor’s advice - that’s great, and I would shake your hand and recommend you to my contacts as the epitome of professionalism. If your defence includes phrases like “I’ve been using ICA since 2000″ then close the door on your way out.
There are even rare cases of vendors who are shining examples of anti-Protocol Passionistas like Chad Sakac of EMC. He even plays nice with NetApp, how’s that for a shining light? I would love to see a Protocol Passionista argue the case for a particular storage protocol with someone like Chad. He should charge for ring-side seats. Another rule of thumb: if you couldn’t defend your protocol faith to someone like Chad, perhaps you’re wrong. Chad isn’t going to flame you if you have a proven method and data to show your choice.
I know you can’t see me right now, but I look something like the picture below (without flat cap) and my message is intended for all you Protocol Passionistas out there: technologies like Nexus and UCS and people like Scott Lowe and Chad Sakac are bringing an end to your world.
14 Comments
Steve – thanks for the kind words, and I think you’re BANG ON. There are such larger architectural/design challenges, we are working to expand the good of block AND the good of NAS (longer term, even converging them via things like pNFS, which decouples presentation from access, so you can have your cake and eat it too).
Converged network designs make the decision a non-decision – it doesn’t cost a penny more (and in fact can save a lot of dough).
Let’s move past the petty stuff, and move on to the new challenges!
I fully agree. For me it’s all about requirements, constraints and risk for a customer. Whatever gets the job done is fine with me!
Great article again Mr Chambers, keep ‘em coming!
Nice post. It definitely is a new world in IT where lots of dogma is being challenged. Boundaries between segments of the market keep coming down as vendors go into adjacent markets or make acquisitions and partnerships to build a complete stack.
Here in the blogosphere, there are few “Protocol Passionistas” left – Mark Farley gave up the war last year (http://www.storagerap.com/2008/10/chuck-i-will-fight-no-more-over-iscsi.html), but Greg Ferro is still anti-FC/FCoE (http://etherealmind.com/tag/fcoe/).
Steve,
Another excellent piece. It’s makes me think about all of the debate in my industry not just around display remoting protocols as you touched on, but also around reference architecture of application and desktop delivery. What immediately comes to mind is Terminal Server vs VDI. I’m so very tired of the countless debates on a topic of which there is no clear answer…no clear winner. Most companies (at least those who actually care about choosing the right architecture) will end up with a mixture of both. Anyone who’s not at least sampling the grass on the other side will soon find themselves out of a job. My .02.
Shawn
Spot on, Shawn. I’m a huge fan of “technology blend” because not only is it the reality in most customers, but it’s the only way for me to make a living as a vendor! If my products can live harmoniously with others and deliver value then I’m out of a job. In fact, blending technologies behind a consistent, brilliant service is the key to customer IT success IMHO and that’s what my Desktops R US post was about. Thanks for the comment :-)
I’ll have to echo Chad and say thanks for the kind words. I’m not so sure I live up to being the “blueprint for the new multi-skilled” architect, but I do appreciate the compliment!
It won’t happen overnight, but the sooner we take leadership roles—such as the one you’re taking here, Steve, to call out this behavior—the sooner we can drive everyone to get over the protocol wars and get down to the business of finding the right solution for problem at hand. Everyone ready?
Great article Steve! I totally agree.
Working for a multi-vendor distributor, I often have to face totally blinded “passionistas”. I can think of so many people that fit exactly the description you’ve given.
Otto
what we gonna do about it, Otto? we need a call to arms…
Scott, you are the most popular example I know of. You know what you’re doing, you’re not getting into virtualization across compute, network and storage because you like spending time away from your family, right? There’s more of us out there, but we are outnumbered. Who else will hold their hand up?
I couldnt agree more. But there is always a tendancy to look after the company who is putting food on ones table, even if its not the right technology for the customer.
There is probably alot of MS employees who have felt this way about VMwares products, and know that it was a way better solution than they offered, yet they can’t go out and say that, or they will be hung and quartered.
As independant consultants though, its criminal not to present all vendor options, and help lead the customer down the path that best suits them. Lets hope this becomes the trend, and it further pushes the vendors to deliver great products.
Cheers,
Jeff
Agreed, Jeff. That’s why I made that point that if you are a vendor you can be exempt from this. Chad is a shining example of someone who transcends the vendor stuff because he can. I can’t. Scott Lowe can because he’s not a vendor.
Perhaps you’ve just coined a new term: It’s criminal to be comfortable? :-D
This is all good and well and any good technologist/architect should not care since each of these protocols have progressed to the point that they are all “good enough” for a variety of situations. One might work better than another in some use case. This is a small part of a full architecture for any company that has any systems currently in place.
Generally in my experience the protocol is the last thing one should be worrying about in a direction / solution to make. Costs to implement the new solutions are a major driving factor. Management tools and in-house experience are nothing to scoff at. If company A is already in bed with Citrix then ICA comes with it. If company B is in bed with VMware View then PCoIP is the protocol. Reimplementing a VDI solution if your using one or the other already is very expensive.
Looking at this form a purely sales position (I can feel the stares already!) I have to agree that I win the majority of my business without mentioning underlying technology until we get to full solution design time. I the lead salesperson for a large reseller in the UK, and while I am predominantly “targeted” will selling VMware solutions, I will sell what fits for each customer. For some, as Ian says, in-house experience is quite a key point for many of the smaller companies that I deal with. They can’t afford to take time out to learn new technology, however, for the larger green-field sites, the decision tends to come from CFO / CEO level, and technology is for the main part irrelevant to them.
As long as the IT teams buy into the concepts that I am evangelising, then my solution design really takes a back seat until we get to the technical verification point.
In terms of the solutions that I’m currently working on, the majority have server, desktop and storage virtualisation in place, and generally speaking a mix of the above-mentioned technologies (iSCSI and FC, RDP/PCoIP and ICA).
Horses for courses as they say, and I apologise for commenting on what to most people appears to be a more Technical aimed board! ;)
It’s really just to prove that some of us sales guys do talk with our techies! Good work again Steve, thanks.
Fantastic post, and definitely right on. Picking a protocol preference does nothing but limit you’re ability to deliver the right solution. Having the ability to choose adds flexibility and lowers cost. When talking about I/O specifically there are several protocols that are the right fit for different applications, why not throw them all down the same pipe and make that pipe big enough and smart enough to handle them? When talking about VDI protocols who really cares if it’s RDP, PCoIP, ICA or IPoAC (see Wikipedia) as long as the VDI solution itself is providing the right performance at the right cost?