 |
|
 |
 |
The reason the human will always be required at some relevant level in the implementation and use of computing technologies is pretty simple - the computer can execute a codified series of events much faster and more consistent than any human, but it can't write a song.
The ability to create - not "think" - is what separates us from the Duo-Core next door. The issue is our society values the Duo more than the Doo-Wop.
I went to see one of my favorite bands, Steely Dan, recently. They are one of the more talented batch of folks you might ever see - from a few (2) guys ability to even conceive of a song to begin with, to being able to layer on many other elements musically which individually might be interesting - but holistically form pure brilliance. I can't see a machine ever writing a song.
So Steely Dan, with it's primary driving creative genius at the helm, had about 10 backing musicians. Each was absolutely amazing in their craft. Each may have been among the top 10 on the planet capable of performing at that level. Each probably made about $50,000 a year. Each could have been replaced by a machine at any moment.
It sucks, but that's kind of like IT (except rarely in IT do I see smoking hot backup singers). The systems we build are superb at executing the tasks we tell them to execute - but those tasks require the creative genius of the strategic human who can "see and hear" what the symphony should be - and we don't value those strategic maestro's of IT in our society - so we end up with well meaning supporting musicians trying to keep up with the whims of guy who owns the place instead of being encouraged to express the creativity we posses. And then we try to automate or outsource those guys too.
IT guys range from the best contract musician in the band to pure road crew. It is very rare that a true IT maestro is allowed to practice their craft - not without lots of "patrons of the arts" being willing to dole out lots of dough simply because they believe it is important. While history is littered with examples of world changing genius, from Mozart to Joe Strummer, why is it there are only a handful of IT types who have been allowed to perform? Beethoven didn't need to develop the piano - he simply mastered its use. The CIO doesn't have to develop the next core - they need to conceive and create a better way to harmoniously use the instruments at their disposal to create magic. The Web 2.0 movement, with all it's ramifications, screams for creative new leadership in IT. It is sort of like we are leaving the Disco and Techno era's - people can't take stupid much longer.
It has happened. How many Google's, Amazon's, MySpace's, etc. do you need to see change the world to see that they were not simply lucky? IT IS the business in the new world order - but you ain't gonna be anything but a hack lounge band if you don't hire a maestro to see it - and then encourage and support them. Maybe an acid trip will loosen you up to the possibilities.
Side Note: You know you are old when the shows you enjoy are the same that you enjoyed 20 years ago - with reverse metrics. 20 Years ago at a Steely Dan show there was much more hair than money, but today it's the exact opposite. It was a sea of 40 something bald men in Gucci shoes. At least there were smoking hot backup singers.
My recent CW ditty follows.
Q: Steve, how can an IT executive expect to bring “strategic value” to the business if the business itself does not value IT, and also shows no real business acumen? P.B. Redondo Beach, CA
A: You can’t.
“Stupid is as stupid does” is about the most accurate statement you’ll ever hear. You can fix most any problem, except stupid.
The bottom line is we are in a vicious circle – IT hasn’t been able historically to prove its true value to the business, and as such is treated as a cost center, which are the first things to lose funding when a business downturn or hiccup happens, which always happen in poorly run companies. There isn’t much you can do about that, except to find a better company who values the strategic possibilities of IT and does so while actually running the top line business better than a six monkeys driving a mini-van juiced up on tequila and Jolt. Easier said than done, unfortunately.
The good news is that for every well run business, there appear to be 10-100 more entities that are either about to succumb to Darwinism or have somehow bypassed the evolutionary process temporarily. Eventually the Dodo became extinct. The better news is that as the natural laws of business are slowly doing their thing, the businesses who withstand the test of time tend to be the ones who have the brains and adaptability to move on to the next plane.
The really good news is that the next plane will be dominated by data. “Gut” instinct will be valued, creativity will be sought, but data will drive decisions. When that happens, IT becomes what it always should have been – the keeper of the crown jewels. I just hope we’re all around when it finally happens.
I’m seeing signs of intelligent life. If you are not up to speed on what the web 2.0 phenomenon really means for you, IT, and business in general, you should start paying attention. I think I’m even going to write a book on it (I’ve threatened that many times before, however, so don’t hold your breath). Web 2.0 is all about pushing new business agenda’s while completely dismantling traditional methodologies for performing corporate functions and turning the economic models of all that we know about technology and infrastructure on its head. There is too much to cover on this topic, but suffice it to say that when the dust settled on web 1.0, the guys left standing have turned the world of IT infrastructure as we have known it into a shambles. What was never possible before is suddenly possible – and what cost a million bucks a year ago might cost less than a tuna fish sandwich next year. All the assumptions that you hold about IT and the business of IT should be put on notice – change is in the air.
Amazon didn’t start by wanting to be in the IT infrastructure business – they had a different idea. YouTube or MySpace sure didn’t desire to be in the infrastructure space either, and nor did Yahoo or even Google – but they all are. The only reason to build your own stuff, if it isn’t the business you are in, is because you can’t buy the stuff you need or you can’t get the functionality you want at the price you are willing to pay. There is value to the video of my kid shoving the cat into the toilet, but how much? Not enough for me to pay YouTube to keep it. Yahoo and Google will give you unlimited space now – gratis. Those models are simple to understand from an IT perspective – they want infinitely dynamically scalable infrastructure with an absurdly low (or no) real cost to acquire or manage. Otherwise, their models fail – and none of those models seem to be failing just yet. Since the entire IT industry is predicated upon the very opposite of this model, it is no wonder that web 2.0 companies had to grow their own ways to provide that functionality at the required economic level. Free means crap, right? Wrong. How many of you keep Yahoo mail running on your blackberrys along with Exchange because you know Exchange will be down sooner or later? When is the last time Google mail was down? So my enterprise caliber application running on enterprise caliber infrastructure with enterprise caliber IT talent is less reliable than my free email? Yep.
What’s most interesting is that the model has been proven internally to these organizations and only is now teaching the rest of us what that means. The email example is perfect – Google or Yahoo are not providing us these tremendous services for zero dollars by simply running enterprise applications on enterprise infrastructure more effectively than we do. They are doing it 100% differently. The question is - who will be the next wave of vendors to take that movement and apply external commercial viability to it. That’s when all hell is going to break loose. I can’t wait.
Now, not to get you all bummed out, but as cool as all this is, there is still way more dumb than smart out there. For example, I have a neighbor that for years has been the CIO of a “direct marketing” company – a glorified pyramid scheme really. This company has been as big as $650-700 million bucks – currently plummeting down to sub $400 on their way to oblivion. They are one of these “party in the house” marketing organizations – the kind we all hate but feel compelled to attend from time to time in order to maintain civility in the neighborhood. Anyhow, these guys sell candles. They have made a ton of dough selling candles over the years by having Marge host a little candle party and getting my wife to stop in and place an order for $37 worth of the smooth smelling wax jars. The issue is that once my wife gets the candles, she likes them. She wants to buy more candles, but she has absolutely no desire to attend another candle party with Marge. She’d like to simply log onto the company's website, and order some more. The company, being stupid, has decided that would be bad for them. They fear that somehow that would break their allure with Marge – or somehow hurt the mystique of the home party. So instead of assigning my wife a customer ID that automatically pays Marge every time my wife orders something (which tends to be a lot), they decide to not allow my wife to order anything on-line. Remember Julia Roberts in the Beverly Hills boutique that looked down on her and lost a fortune in sales in Pretty Woman? “Big mistake.” My wife doesn’t care. She goes to the mall, CVS, or any other on-line store and buys candles. Marge is out a commission, and the company is out a customer. By not only ignoring the evolution of their own market dynamics, but flagrantly disregarding basic customer desire, they are destined for the scrap heap. What did they do when they started to lose customers and piles of money? Did they mine their data to find the root of the problem? Nope, they slashed costs, killed every IT project, fired developers, and made sure the only folks left were political allies without a clue.
I love Darwinism. Keep the questions coming. Ciao.
Lets assume we are Joe IT guy in the ABC company – an upper middle market company with a few thousand employees, a dozen sites, and all the problems folks like us deal with. We run our transactional production systems and our distributed windows stuff. We have big SANs and file servers. We have stuff everywhere. We back things up, we do some DR. We are tragically overworked and chronically undervalued. We are Joe IT.
Let’s pick one little area where process improvement can yield big results – Test and Development. Everyone has T/D operations. Most operate in some version of the following:
- T/D is used to make sure internal and external software applications, new infrastructure, and upgrades work the way the are supposed to prior to them being rolled out into production.
- In order to perform step 1, T/D has to get real life data that is complete and current from production systems into their own systems.
- In order to perform step 2, T/D has to beg, borrow and steal – and lives at the mercy of the production people. It takes time, planning, and prayer. The application normally has to come down, the database quiesced, the infrastructure specialists turning knobs and pushing buttons, and the data moved.
- Once step 2 is complete, the actual testing can occur. Usually T/D will make additional copies of the data sets (which are probably already out of date by the time they are used) to test different things.
Now, let’s talk about the practical inefficiencies of what just happened. First, the production system runs some application(s), most likely on top of a database. We tend to find our production data really important, so normally we would have at least two copies of that data. We tend to keep our production data on very expensive, very big, very power hungry infrastructure. We create the data on that infrastructure, and then we keep it there. We then make more copies of that data, also tending to keep it there. We might have 2, 3, 4 + copies of the exact same data, at various points in time or at the same point in time. That means our cost of housing that data is 2, 3, or 4 plus times the cost of housing it as a single instance – and not just for capital – but for power, cooling, footprint, etc.
When we make our test copies, we also tend to create and keep those copies on our production system. Sometimes we delete them when we are done, but sometimes we leave them around for a long time. Sometimes we forget about them altogether. Those copies take up more space, power, and brains. Those tend to be backed up along with everything else on that mission critical production system. That means we might be backing up 23 copies of the exact same data in one single backup. If we do a full backup each week, we will create new backups of the exact same 23 copies of the exact same data each time. You see where I’m going with this?
We have test servers that we run that sit around sucking juice whether they are used or not.
It gets worse. Not only was the process of getting the database copies difficult, and expensive, but no one considers the security implications associated with having potentially hundreds of copies of real production data floating around – in the production system, in the test systems, on the backup systems, at the DR site, and on the tape at Iron Mountain. Backup is good, right?
Having 2, 10, or 100 copies of the same, non-changing data sitting on “production” systems isn’t even the real problem. The downstream effects are the issue. Extra stuff on the production system slows down the production application, slows down the database, and slows down the user. We tend to combat this by buying more hardware. Bigger, faster hardware that sucks more juice, takes up more room, and causes more disruption. More “data” means other processes suffer – networks get clogged so we need more bandwidth, backup servers get bogged down so we need bigger machines, backup targets get full faster, etc. The rest of our processes don’t know or care that it’s the same data – only that there is more. More causes problems. Problems cost money.
The conundrum is that if you are the vendor in the production system, you kind of like it when Joe IT calls for more stuff. It’s hard to tell Joe not to buy more – but instead to just change some of the behavior that causes the problems. If we were to try to really help Joe, however, we’d lay it out like this;
- Define the objectives for Test and Development – in a perfect world
a. Get a complete, accurate, and timely copy of the exact production database
i. Zero impact – non-disruptive to production
ii. Non-disruptive to production IT
iii. Automated
b. Put that copy somewhere else – NOT on the production system
i. Don’t create work for production systems or people
ii. Do it dynamically
iii. Run virtual machines everywhere you can
c. Create a protection policy for the Test/Dev data
i. Do we back it up?
ii. If so, when, why, and how often?
d. Create a security policy for the Test/Dev data
i. Protect the assets as if it were still in production
1. This assumes you are not TJX or TSA
ii. Enforce disposition/destruction
e. Define a Data Repurpose Policy
i. Who else could use a copy of this data?
1. Should we use this copy as a backup copy?
2. Should we replicate this copy as a DR copy?
ii. Are there other applications that could use this?
1. Data Warehouse
2. Business Intelligence
3. Those guys in marketing
4. Business partners
If Joe did this, nice things happen. By simply grooming the production systems of Test/Dev copies, Joe unclogs a lot of space, performance, and all the other associated costs. Joe just took a universal pain in the rump from the production IT staff and made it disappear. The Test/Dev folks have a consistent schedule to get fresh whole data sets to play with. The security people are happy that there are less copies of stuff everywhere. The backup guy is happy that the extra stuff isn’t killing his systems. The Finance guy is happy because the cost of the test/dev infrastructure is an order of magnitude lower than the production stuff. The licensing costs alone for running the applications and database on bigger machines are huge – and now they will be pushed our or even go away.
If Joe did this, he might also figure out that not only should ABC keep Test/Dev data off of production – but that even the production data itself is “groomable”. 90% of the data that makes up production is fixed content, or data that isn’t going to change. It is no longer “dynamic”, and as such, the same questions can be asked of it. If it isn’t going to change, should we still back it up the same way? Should we still have 4 primary copies of it? Shouldn’t we put it on infrastructure that has attributes that are more aligned with the current state of that data?
If the data isn’t changing, but the attributes are – then continuing to do things the same way is illogical. If, after some period of time, access of a certain piece of “formerly transactional” data went from frequent to never, why don’t we put it through the same exercise that we did with the Test/Dev data? If we moved that data out of “production” physically, but maintained its place logically (i.e. we could still access it in the same way if required – but it might take longer to show up) – we could then do some really interesting things. By delineating our production data into Dynamic and Static (fixed) buckets, under a single consistent logical view, we would change things forever. Our static data would have a different “lifecycle” – with different attribute requirements as time (or whatever metric) moved on. We would move it out of our “tier 1” outrageously expensive gear and onto much lower cost gear, which would mean that our tier 1 gear would perform at optimal levels all of the time. In theory, if our Dynamic to Static flow was predictable, we may never have to buy another piece of hardware or upgrade another license for our production systems again. We would not only move data off of our most expensive, mission critical stuff, but we would change the attribute requirements of that data when we did. For example, we might keep 4 primary mirrors of that data while it is dynamic, and back those up every hour to a disk target, and replicate those targets every 8 hours off site, etc. – but we’d stop doing that once the data became static and moved to the “static production” state. Perhaps we would make sure we had one mirror locally, and one on-line backup locally, and another copy at the DR site, with 3 “oh no” copies on tape. Then we’d never back it up again – because what would the point be?
If you did that you’d save a lot more than 25% of your power and cooling budget. You would save so much money you would probably affect the earnings per share of your stock. You might want to negotiate a bonus up front!
Vendors, who naturally are opposed to such intelligence unless they aren’t the ones getting all of your money, are starting to wake up. It’s hard to upset the gravy train, but they are starting. The fact is, the very nature of data is changing, so just jamming more of the same down your throat might help them short term, but eventually will cause them more problems because you won’t be able to ingest any more stuff from them. They are already seeing it. Those that learn to cannibalize themselves are the ones who survive long term. The principal of adaptability in Darwinism works.
IBM bought Softek to do things like this – at least some of the things like this. They are the ones who are at the very core – they are the Mainframe. By providing customers a way to groom production systems, they take food out of their own mouths. There have been heated arguments inside the blue machine for just this reason. The bet is that while they may push off mainframe revenues, they pick it up in other areas. EMC owns tons of the storage market attached to those mainframes. Oracle gets a big chunk of that pie too. By helping customers groom their production worlds and create new efficiencies, IBM takes money away from EMC and Oracle – and gives themselves a shot to compete for that storage business, and the data management functions that Oracle provided. Network appliance has been working with Solix to pull things out of those same environments and have them land on Netapp boxes. Not only is a Netapp box a heck of a lot less money than a big giant tier-1 mega array, but once the data is there people are astounded at how easy it becomes to do really useful things. One big wig at a Fortune 1000 told me that they could now create zero-footprint copies of their production data instantly using Netapp’s writable snapshot, which lets them do things that haven’t even been remotely possible previously. He said “it was like we hired 4 more people with the time we saved – and now things that took days take seconds.” Most folks really can’t afford to stop production to do backups or run reports, but they have had no choice. Now they do– and they now do it securely. EMC won’t like losing the core disk business, but if they can take the server dollars, the application/database dollars and get the customer to see the light, they have a whole new opportunity to sell the other thousand things they have in their bag. If it’s fixed content, why not let Documentum manage it? Who needs Oracle? It’s sort of like being a stock trader – you don’t care whether it’s up or down, only that a transaction is occurring. If you are an “emerging” player – Isilon, Pillar, or anybody else – this is an opportunity to shine. If you are a little guy trying to displace NetBackup, this is your chance. Real change opens doors.
Process changes of this magnitude are truly game changing – for IT and vendors alike. It tends to take a massive event to get either side to move out of their comfort zone however, and not being able to buy any more power might just be that event. Once something like that occurs, and people are forced to look at new ways of doing things, and then Pandora’s Box might fly open. With the lights turned on, people can see. Sometimes things don’t look as good with the lights on – just ask my wife.
In last month's Storage Magazine article I explained my theory of why we are so screwed up infrastructure-wise, or at least how we got to this point. Now I'll give you the way out. Then, I'll talk about cars.
Forget everything you know for a few minutes, or at least everything you think you know. Accept my argument that most everything we’ve done in commercial IT has been done based on transactional requirements. Open your mind. Breathe through your eyes, Danny.
There are two distinct types of data – dynamic and persistent. Dynamic is data that is in flux – this is where transactional data begins. Persistent data is just that – fixed. It doesn’t change. It is what it is, and will never be anything else.
Just because data is dynamic doesn’t mean it starts and dies within an RDBMS. Structured database data certainly starts as dynamic – but at some point it becomes a non-changing record. It’s persistent. You may have reasons to keep it inside a database forever (although I doubt they’re valid ones), but those records are still persistent – they are what they are.
Rule #1: Don’t confuse how something begins its life with how it will end. Everything begins dynamic and ends persistent. Stop delineating between structured, unstructured and semi-structured. All types live dynamically for some period, whether it’s a Word document, a movie, a credit card transaction or an email. It all ends up as fixed digital content.
Rule #2: The attributes and requirements for each type of data are vastly different – in most every sense. Read/write performance, throughput, redundancy, protection, disaster recovery, etc., count more in the dynamic phase of data life – but we’ve extended all of those philosophies to data that stopped changing, which causes 90% of our operational issues. Building data redundancies and protection schema’s to handle real money transactions is good business – backing up a non-changing data element a thousand times isn’t. Keeping your bullet proof transaction system capable of handling all the dynamic money events thrown at it is good business. But adding processing power, capacity, network infrastructure, etc., to keep it churning away rather than removing the 90% of the data that isn’t dynamic or possibly even relevant that can interfere with the real transactional stuff isn’t. Thinking you can’t make this happen is wrong.
Rule #3: The ratio of true dynamic data (and data being “treated” dynamically) to persistent data is about 1:10 – and that ratio will rapidly evolve to 1:100 and beyond. Dynamic data just doesn’t stay dynamic for very long.
Transactionally oriented systems are all about doing things fast. Perform the transaction fast, store the data fast, load the data into other systems fast. If it sits in a database, it’s easy to find – which is the point of the database. If we could, we’d have everything inside a database, but they weren’t designed for that. The persistent data world is all about finding things. That means you need to know where to look. Having stuff everywhere, inside and outside of the building, makes this task all but impossible. The whole categorizing/classifying/indexing/search thing going on today is designed to add structure so we can find things. It just seems to me that if we created two distinct “virtual” places to look for each distinct type of data, it would be a heck of a lot easier to find what we want in either. If all our dynamic data sat in one place designed to handle things like that – no matter if it’s Word or big dough – and then was moved (based on business rules) into the persistent digital content store, we’d be able to architect this store entirely differently than the dynamic store.
If the dynamic store is about speed and redundancy, then the persistent store is about infinite dynamic scale, the ability to find things easily and quickly, autonomous self-managing/self-healing infrastructure--and it should be really, really cheap to buy. If the dynamic store requires knob-turning specialists, the persistent store requires the occasional dusting. Stop trying to turn the dynamic store into the persistent one – and stop trying to make the persistent store dynamic. If you think differently, you’ll act differently. If you act differently, you’ll realize you can get back to making IT a competitive advantage.
And Now For Something Completely Different....
If I had stupid money, I'd buy a Mercedes SL65. The car wouldn't get noticed at the shopping mall, but it is a 700HP rocket ship. Silly, really. It is the very same car my girlfriend, Lindsay, just smashed while being smashed, out in L.A.
I stumbled across this thing called "Supercar Life" (www.supercarlife.com) in one of the high end rich folk magazines I read and drool over. It turned out (in a very legal, tax deductible way) that one of my biggest customers is a car freak, so as a purely business oriented gesture, I took him to this. The owner, Joel, a guy with way too much dough on his hands, figured out that there are a lot of "men" who still act like they are 12 when it comes to cars. So, for a piddly $5,000 you can fly to Pittsburgh where he rents an entire race track (Beaverun), and only lets 10 folks on it. He has (2) of each - brand spanking new - Ferrari F430, Lambo Gallardo, Porsche 911 Turbo, Astin Martin DB9, and the SL65 (Joel had his $450,000 Porsche Carrera GT sitting out front, but didn't offer to let me drive it for some reason). There are about 6 or 8 real race car driver guys - not local car club pro's - real dudes that race really expensive cars really fast that teach you how to not destroy the very nice cars that you don't own. Every few minutes, you swap cars. By the end of the day, you have driven a million bucks worth of cars really, really fast. It was awesome. If you are thinking of buying one of these rocket ships, why not spend an entire day test driving them first, and learning how not to "Lindsay" your ride? It was really first class - they put you up at the Marriott Resort and Coal Mine, feed you well, and were just terrific folks. My customer may have had to have surgery to remove the smile from his face. Now I just need to find some other car crazy customer who gives me piles of dough to justify my return trip.....
This is a recent article I wrote somewhere.
We’re hung up on an outdated computing model that makes everything tougher.
We’re at one of those rare, very powerful inflection points in our crazy storage universe that’s so big it can’t be ignored, but came so slowly that nobody seemed to notice it until now.
For 50 years in commercial computing, dynamic, transactional data is what moved computing out of science labs and into the commercial world. Trillions of dollars have been spent trying to harness the power of that data and provide the infrastructure to create, store, protect, move and manipulate that information. The problem is that most data isn’t transactional anymore; a huge portion of the issues in storage and throughout our infrastructure are caused by us trying to use the same systems, architectures and methodologies we’re used to for data that has new requirements.
Commercial computing was built on the application of technology to support business causes—to create competitive advantages by enabling users to work more efficiently. That led to greater profitability and faster decision-making based on having more data analysis. The competitive advantage eroded when everyone had computers, and went from being strategic to being tactical and defensive. The game became about how many transactions you could support concurrently, and the value of each transaction minus the cost of performing those transactions equaled the “hard dollar” profit.
The industry built big iron that cost a ton, but it didn’t matter because it was easy to articulate cost vs. profit. Companies spent the next 35 years trying to outdo each other with the number of transactions they could process. Distributed computing lowered capital costs and raised operating costs, and things became more nebulous. EMC convinced users that storage was as important as the server, creating a new market that it has since dominated. Other vendors have come along, but most offer cheaper or better mousetraps still designed to deal with the fundamental transactional nature of data.
Commercial computing was dominated by folks who built the right stuff at the right time and sold it most effectively. IBM mainframes are still the transactional systems of choice. Oracle is still the dominant DBMS. EMC is still the core storage device. Hewlett-Packard and Sun Microsystems cashed in on the rush to distributed computing. Smaller servers meant smaller decentralized storage, opening the door to the likes of Network Appliance, which created and continues to dominate the file-serving world created by the distributed computing evolution. Everyone and their brother came out with a block array for the “open” world.
What that means is that we now find ourselves with massive amounts of things. The number of devices we have to manage in complex networks rivals the human nervous system. And it’s compounded by our applying the same basic operational and technological constructs to this world as we did with the monolithic systems of old.
Industry used a new market wave by dumbing down and cheapening the heavy-duty stuff you’d already bought. They took systems built for transactional business, ripped stuff out to make them cheaper and then sold them as the answer for the “new world.”
We now have 100 times the specialization we did in the centralized world because we have 1,000 times the amount of things to know. We do this because we haven’t noticed that we’re solving the wrong problems the wrong way … and if we don’t stop, it’s going to kill us. Next month, I’ll explain the right problem, which will get you looking for the right answers.
Remind me never to complain about being in Orlando again. The weather has been stellar, and back in Boston it is 35 degrees and raining a foot an hour. Glad I'm not a marathon runner also (I'm not, though that may surprise you), as they discussed canceling the event for the first time ever. They elected to go forward, which is ridiculous, but should make for entertaining video on the news. Why I continue to reside in the great Northeast is becoming a better question all the time. Arizona seems nice.
Orlando must the U.S. leader in cheesy gift shops, surpassing Vegas in my humble opinion. Middle America is frightening. It is hard to imagine ourselves as the tech whizzes and global super power extraordinaire when in line at a drug store behind four non-related (seemingly) people with full-on mullets, and supremely bad facial hair - and those were the ladies.
I've already begun to receive the usual bulk batch of press releases since SNW starts today. Emulex announced an independent company validated their performance is superior to Q-logic in a VMware environment. The only problem is I've never heard of the company (Demartek) and the full report link doesn't work, not that I'm insinuating anything. So yes, of course I am biased in such things, I don't think anything worth proving should be done by anyone other than ESG Labs, so take it with a grain of salt. Having said that, testing aside, the play is brilliant. Why wouldn't you attach yourself to VMware in any way you could? I'm surprised that more folks haven't figured that out - creating parallel branding to VMware is a very, very smart thing to do right now. It will be interesting to see how many of the vendors at SNW have overt VMware programs under way.
In this morning's USA Today, the front page banner is "25 Stocks You Should Have Bought". People pay for this? Why not pay the hotel staff to sneak into my room and hit me in the head with a hammer instead?
Speaking of VMware, the story they tell is one of the best you'll hear. I'm having trouble figuring out how Microsoft or Xen, etc. will be able to catch up. This game may be over before it starts and others will have to figure out how to play above (or below) that line. Second, EMC really has to be commended for leaving VMware alone - though now I think they have no choice. I'm amazed at how separated they really are - you would think that there would be much tighter integration with the EMC product set, but if anything, it's the opposite so far. EMC might be being a bit too nice in all of this - words I never thought I'd say.
The Dutch, Data Domain is also run by a Dutchmen - Frank Slootman. And yes, he is also tall, and about to be able to buy a lot more Amstel's. DD is on the path to their IPO which could happen in short order. At the market caps of other recent tech IPO metrics such as Riverbed, Isilon, and Commvault, It will not surprise me that DD comes out big. They sell a lot of stuff, and have the luxury of really being the first and broadest in the whole data de-duplication space - which today is used primarily for backup targets but that kind of technology will have a play in all aspects of storage. They settled their differences with Quantum over some patents, so I think it's show time.
Speaking of IPO's - has anyone ever quantified the business impact of a company pre-public and post? It makes sense that prospects on the fence about buying from a smaller private company suddenly feel their investment is worthy if a company is now public, but I wonder what happens to things like the average length of a deal. I assume it becomes a shorter time from prospect to customer, but I've not seen any data on that. I guess I'm wondering what other real business value is derived from the IPO exercise, besides the obvious. What I do know is that once you are public, life doesn't get easier if you are the CEO and management team. If anything, the pressure becomes far greater and forget about privacy. Being a mucky muck has always been hard, but being one with all the scrutiny that comes from being public is whole different ball game. It would be interesting to analyze data on how CEO's have faired in the post Sarbanes-Oxley era of going public. I bet it's a good time to buy Mylanta stock.
Speaking of stocks - has anyone paid attention to the bizarre coincidence that EMC's stock is up about 10% since there has been a bunch of press about how Joe Tucci and others are paid? Joe's good, but not good enough to suddenly make the price of the stock go up just for the heck of it. Most of his compensation is already tied to stock appreciation, so I'm pretty sure he's on board with the whole "raise the stock price" issue. It doesn't seem reasonable for shareholders to be able to vote on compensation, because only a B player CEO would ever take a job in that circumstance. I want a CEO making a fortune if the compensation is tied to the right metrics - like increasing shareholder value. The last thing I want is some nitwit dumb enough to take a job where stockholders and employees get to vote on their pay. There is a reason why big successful companies are what they are, and to think you can get a quality chief on the cheap is just plain dumb.
Last week I went off to do a speech for Network Appliance in Amsterdam. I used to do this kind of stuff all the time - the money is great and I get to go to really super places - but I rarely do it anymore. After a million flights, I found that a few years ago I became a white knuckle flyer - the kind normal flyer's hate to have on their aircraft. Doctors and shrinks tell me its natural, after my bout with cancer, but I don't think that's it. Anyway, I see no point in interjecting fear and terror into my own life, after all, I have a 15 year old daughter to do that for me.
This time I said yes, partly because I really needed a few days out of town and partly because, well, it's Amsterdam. I took my lovely wife, Jess along, figuring I couldn't get in that much trouble if she was with me. A six-foot blond tends to draw attention away from me in most situations.
I have several observations; first, the Dutch are giants. Being statistically average in many ways (5 foot 7 inches is exactly statistically average, though tall people disagree) I never really stand out in a crowd. In Amsterdam I was an oddity. Every Dutch male was at least 6'3". Jess was average - words I never thought I'd say. Everywhere you looked, there were tall, gangly Dutch folk peddling away on their 1943 looking bicycles. All of them looking at me and seeming to say "Look Hans, a little person..."
The Dutch ride bikes. Lots and lots of bikes. Unlike Beijing, or Rome, where bikes and scooters also outnumber cars, in Amsterdam the cohabitation of bikes and autos seemed to work. I couldn't figure out the rules, but it wasn't "just go" as it seems is the rule in the other spots. I didn't see any flattened Dutch while there.
I took the red eye to Amsterdam, landed, got changed at my hotel, and went to the event. My wife went to sleep. I enjoyed a fine lunch with a handful of Netapp customers - two of which stood out; a gentleman from Phillips asked the most intelligent questions of my Dutch host around the areas of data center virtualization and another from @Home, who apparently either didn't go out of business in the bust as I thought, or no one told the Dutch. The Phillips fellow knew that adding virtualization for utilization and operational benefit at the server level was mandatory - but recognized that once that occurs the ability to see end-to-end from a virtual server through virtual storage was all but impossible. I couldn't agree more. The @Home dude was explaining how Holland is a wired country - effectively the Verizon FIOS equivalent of the Dutch - and that they were able to pump outrageous levels of bandwidth direct to the household. (I'm dying for FIOS, but of course can't get it. They have to dig up my street, which means it will be years. My buddy has it a mile away from me, and thinks that now that he gets HBO he is a power user. My kid is playing interactive video games with others around the globe while I pipe video all over the house through my MCE and am dying to get into IPTV, but I have to wait.)
My speech was entitled "How Transactional Systems Are Killing IT" - which is a darned good one if you ask me, though I'm not sure how my logic translates. I'm also quite certain I was ridiculed by the hired Dutch MC. (I know he was hired due to his outrageous paisley jacket). My premise, which I'll explain in my next Computerworld article, is that our entire industry is based on the commercial computing success spawned by transactional systems, and that ever since then we've tried to come up with more and more products based on that same architectural principal. It was OK until lately, because now the nature of data itself has changed. What was valued upon being dynamic or transactional is now static or fixed digital content, and yet we continue to attempt to apply the same methods as always in dealing with it. It won't work.
I then had a Heineken and returned to my hotel, to find my wife all dolled up and ready for dinner. We were invited to attend a small dinner with some Netapp folks and their local channel partners. We were graced with the Big Gouda, Dan Warmenhoven (oh yes, a name as Dutch as they come) and his lovely wife. Suffice it to say the poor channel folks didn't get to say too much. (However, it is important to note that one of the very tall Dutchmen at the table had apparently spent a significant amount of time attempting to pick up my wife right in front of me, which of course I didn't notice. I don't think it was until the cancer story that he really gave up.) I had a great time applying sports metaphors to business concepts with Dan while our wives discussed everything from the Red Light District to tulips, and the channel guys tried to figure out just what was going on.
The following day Jess and I goofed around, saw all that was to be seen, and then we left the country before I got into any real trouble. I can't wait to go back.
This appears in the current Storage Magazine
The British are coming!
Good old American ingenuity? Not quite; the Europeans might just be eating our technological lunch.
Actually, it’s all of Europe that’s coming, not just my U.K. friends. There’s a tidal wave of activity across the pond these days and, for a change, the Europeans are leading the way in a lot of areas.
First, let’s look at why it’s suddenly attractive to be in Belgium opening a tech company when previously the only reason to be there was your fetish for beer made by monks. The exit money in tech startups has only really been on our shores—mostly in the NASDAQ. The fact that the market tanked and venture capitalists had to get second jobs at the local
retail store meant there was no real inherent value the U.S. brought to entrepreneurs other than perhaps a larger talent pool to pull from. That excuse has drawn to a close with the
Europeans spending huge amounts of money teaching their young how to acquire the skills necessary to compete in the technology world. They slowly began offering U.S. companies outsourced manufacturing services a few years ago—for example, Ireland gave U.S. companies tax-free status if they manufactured over there—and in the meantime provided valuable jobs to a whole lot of folks. Numerous other countries followed suit. In a way, what the U.S. did for the European Union in the technology space is just what we did for the Japanese in the auto industry—we showed them how to beat us at our own game and then became their best consumer.
Better yet, the Europeans aren’t content to just create the next Silicon Valley (Vallée du Silicium, Silikontal, Valle del Silicio, etc.); they’re leading the business advancements of IT as well. Compliance issues aren’t new to the Europeans. They’re way out in front of most of the world when it comes to dealing with privacy issues, operational risk management, IT service management (ITIL) and even the green initiatives (ROHS). We think Yanks invented the power/packaging/ cooling issues that are so hot (literally) today, but anyone with a data center operation at Canary Wharf knows exactly how to measure metrics from performance to TCO based on one square meter of data center space. Oil crisis? We didn’t see it coming in 1972 and we didn’t see it coming last year. Did you know it now costs more to have your PC sit and run idly for three years than it does to buy it? The Europeans do.
Hermes SoftLab is in Slovenia (it developed OpenView OmniBack, which is Hewlett-Packard’s data protection software). Clearpace, CopperEye, Njini and Xyratex all have offices in the U.K. Caringo (the guys who started FilePool, which became EMC’s Centera) is in Belgium. Fujitsu Siemens Computers sells approximately $800 million worth of stuff every year, but most of us have never heard of them; these guys are in the super-hot virtual tape library market with a unified mainframe and open-systems product called CentricStor that has some of the largest European shops almost crying with glee.
Kaspersky Lab is a Russian security player. I don’t know about you, but when I hear someone say “Russian security,” I figure “How can it get any better than that?”
So it’s time to open up your minds and look to the east if you want to get a feel for what the next big thing is going to be. Oh, and I hope you like cheese with that disk array.
My latest in CW on thinking about power, cooling, air flow, standards, and all sorts of other fun stuff.
I loved Chuck Hollis' (EMC) blog, and equally enjoyed both Dave Hitz and Tony Asaro's responses. With EMC, Netapp, and another ESG'er pontificating on an issue, it must be good.
All of the aforementioned folks are smarter than I, but let me give you some different things to consider.
First, I'm an iSCSI proponent. I remember meeting an intern who worked for Andy Bechtolsheim (who's name I can't recall, but I do remember he looked 12 and had more brain capacity in his pocket protector than I've ever possessed) who was charged with figuring out how to push block data over IP. I think it took me 13 seconds to recognize the implications. It took me a bottle of wine to come back to reality - which is: 1. even if you could use IP networks for block traffic, it will be a long time before people use it to replace core data center technologies, such as fibre channel and 2. eventually it will not only replace fibre channel, but whatever other new specialized connectivity protocol/transport the world comes up with. I shall explain.
I remember within a very short time of the iSCSI initiative announcements people immediately got on one side or the other - there was not middle ground. I was in Silicon Valley when Nishan announced their IP storage router (in an awesome display of bravado, arrogance, and downright stupidity) by telling all that the days of fibre channel were over. They said this in the height of go go valley boom, when there were more than a dozen fibre channel players all getting piles of dough and driving Lamborghini's. My timing was brilliant, albeit pure luck. My cell phone rang with abandon, and everyone from the Wall St. Journal to the Financial Times was asking what I thought about Brocade going out of business. I remember Ashok Kumar from Piper Jaffray getting on CNBC or some equivalent and discussing the end of the fibre channel market. I believe it was early 2000.
Billions of dollars of value were traded in short order. I got a call from Brocade and was asked to come in. I had never met Greg Reyes, but knew most of his team, and certainly knew of him. They wanted my take - and I gave it to them - death is not near, but it is inevitable. I haven't changed my tune. Greg told me point blank "you can't put block data over IP". I suggested he not say that. I suggested he instead say "Brocade builds block storage networking equipment designed to enable sharing of high-performance, mission critical storage assets across multiple servers. In order to guarantee both performance and more importantly - reliability - we do that today using fibre channel. We don't care about fibre channel per se, nor any other specific protocol or transport - we care about providing the most reliable block networking services available to the thousands of customers we have around the world. If IP, or any other technology advancement comes along - as things always do - Brocade will adopt them and offer them to our customer base." In short, I suggested he tell the world this was no big deal, and if it mattered eventually, they would offer it.
He didn't. He said "you can't do block data over IP, you simply can't". He was wrong. The message was wrong, and this single statement did more to polarize the factions than anything else in this debate. (Brocade, of course, has subsequently adopted this position, rightfully so, as have most all fibre channel players. Once you remove the emotion and realize no one is throwing out the baby with the bathwater, cooler heads prevail.)
So Chuck's observations are hard to argue with - and Tony's right in stating that it's really a matter of perspective. Since I prefer my perspective, and feel I'm the most qualified to share it simply because I was there on day one (with no regard for competence, mind you), here's what is going to happen and why.
History is littered with proprietary solutions that are developed to solve very specific tactical problems that turn into markets which last long enough for some to get rich, others to die trying, and then Darwinism does its thing and some sense of reality is applied to that market for that period of time - and to the victors go the spoils. Then, eventually, the reason for that market no longer will exist - it will be replaced with the next thing, be commoditized and standardized, and the original proprietary solution will cling to life as long as possible with arguments like "if it ain't broke, don't fix it". Water always finds the lowest point, and so do technologies. Remember MassBus (I know, I'm old.), SDI (then ESDI), IPI and a zillion other block interconnects and protocols which existed in the day - only to ultimately give in to SCSI for a protocol (and a transport) which won because commoditization and Darwinism do really work. How about Token Ring, DECNET, Banyan, x400, blah, blah, blah. IP won. Ethernet won.
Things don't eventually win because they are better in the short term - they win because they are more ubiquitous. They become more ubiquitous because they take advantage of commoditization economically and VALUE can be generated because of that shift. IP killed the other networking technologies over time. Once we gave in, life didn't end - because only those who fought IP died. Those who recognized the inevitable were able to plan accordingly and either jump on the bandwagon and fight to own the commodity market or moved their value up the food chain above the commodity fray. Those who hoped it would go back to the old days or defended the fact that they were better, slowly and painfully died. Banyan might have been better, but now they give me phone numbers on-line. (Switchboard.com).
The reason iSCSI wins long term is because the ubiquitous market elements are in place (IP is everywhere and Microsoft has added it to the O.S.), and the cost is zip. Betting against that is like being the guy at the party talking to the prettiest girl when Tom Brady and Derek Jeter walk in and still thinking you are the guy she's going home with. Theoretically possible, but I don't like the odds.
So, since you all now totally agree that I'm right (eventually), from a more pragmatic perspective, here's why you shouldn't care. 99% of the time today, iSCSI should not be used to replace fibre channel. It should be used to augment and enhance all the brilliant things fibre channel does. 90% of the servers in fibre channel SAN shops are not connected to the fibre channel SAN. If the block storage network is good enough for 10% why isn't it good enough for the rest? iSCSI enables companies - from small to enormous - to extend the value they received by having a FC SAN to their internal mass server/connectivity market.
What surprises me is that the EMC's/IBM's/HDS guys who are in the core haven't been the ones leading the push toward iSCSI. The name of the game is "filling up the box" and what better way to help a customer spend more money with you than to show them how they can take control over the data on 90% of the servers vs. 10%? To NetApp's credit, they did figure that out a while ago. They ship a huge percentage of their Filers with iSCSI connectivity as well - though I don't know how many use it nor what percentage of the systems are used for block vs. file. Regardless, the other guys own the core, and if they don't use iSCSI to add connectivity and capacity to the core (which steals share from all their downstream competitors, fyi), it's naive to think that someone else isn't going to create a parallel iSCSI network. Once that happens, users will realize that this stuff is pretty good, and as a matter of fact, some of the "cheap" arrays I bought for my cheap iSCSI SAN happen to work way better than my more expensive core stuff. Oh oh. Do you really believe none of Equallogics' thousands of customers are big enterprises - or better yet, big EMC enterprises?
The argument that the SAN guy doesn't want to deal with it is also inevitably destined to fail. The backup guys still don't like to deal with desktops and laptops - and hate the thought of having to deal with remote office data - because their world is held together with spit and chewing gum. Adding more complexity to their lives is not something they are interested in - but it doesn't matter. Whether its mandated by law or becomes the good corporate practice it should, the market will force those attitudes to change, as it will with iSCSI.
The one point I think Chuck nailed was the services piece, but not around requiring services to move from FC to iSCSI. Rather, the services - which will ultimately lead to automation technology opportunities, lie in moving from DAS to SAN - or DAS to NAS - regardless of protocol or transport. No one seems to be arguing that networked storage and disaggregated components is not better than tightly coupled "systems" at least.
So, the good news is iSCSI is righteous, and it is happening. The bad news is that the society we live in (IT) would benefit greatly by a more rapid adoption of this, and many other commoditized/standardized technology implementations but tread slowly because they aren't being shown the strategic potential they could gain by doing so by their big vendors - who have a natural vested interest in selling more of the same stuff they know how to make, sell, and support.
Darwinism always works, but it takes a while sometimes.
Jonny Oltsik's latest is superb - and right on.
The best part is we've read this book before - I especially like his comparison of IPv6 to Y2K - which bodes well for the likes of Cisco above all - there will be a huge amount of money spent on mandated upgrades. Did anyone make more money than IBM in the Y2K doom and gloom?
Second, his point about how absurdly over-funded startups have become marginalized as features inside bigger fish is something the storage business has dealt with for years. CDP anyone?
Best of all, I love the prediction where public enemy number one Microsoft ends up winning - again. All is fair in love and computing, but we spend way too much time focusing on wrong issues predicated on emotion than actually trying to win by fixing problems. Crapping on MS security is easy - of course they are the primary target - they sit everywhere. If you wanna be a real criminal you go where the money is, not the donuts. If you aren't part of the solution, you are just whining. Most of all, I dig the fact that they are going to prove me right sooner rather than later for a change - encryption is the only way to really protect information itself, and they will make it commonplace, erasing the voodoo mystique and unfounded fear of the unknown. Will it screw up? Of course, but give them the credit they deserve, once they start, it's only a matter of time.
This is my most recent Storage Bin. Interestingly enough both CommVault and Riverbed are rocking since thier IPO's - market caps up, up, up - and both are acquiring new customers and revenues at amazing clips. The article, via the process of editing didn't do my NJ Backup friends justice as they have been profitable for a while now (the word barely seems to have been ommitted).
DoubleTake is priced and set for tommorow, and Isilon looks like next week.
Keep 'em coming.
Oh good, we're still stupid
And with nouveau-billionaire companies that are still looking to turn a profit, it doesn't look like we're going to get smart real soon.
"Can you believe my board is actually complaining that we're profitable?" This was the question I was asked over dinner with a storage industry CEO last week. I can't name him because there are forces on his board who, while clearly moronic, may be able to read.
Unfortunately, yes, I can believe it. History does repeat itself, sometimes right after it's finished. We seem to have learned nothing from the lessons imposed on us during the stock market crash of 2000. We were decimated financially because we bought stock in Moron.com when it consisted of five 16-year-old skateboarders, a PC and a market cap that surpassed the gross national product of Asia. Little old ladies were playing bridge and talking about Amazon's IPO. And hard-working people were lured into the frenzy by dirty, rotten stock analysts and ended up losing their retirement savings.
It was only six years ago folks, not 60. Remember buying into those cute Pets.com TV ads? Don't you remember wondering how it could be successful the first time you went to buy your dog food online only to find out that the $2 you saved was wiped out by $27 worth of shipping charges? What about the promise of the original Furniture.com? Great idea, except you might like to see how your butt fits into a recliner before you buy it. Have you forgotten that most of the hysteria defied all logic and common sense?
I didn't. It was stupid. It forced those companies that survived to become real businesses--ones that made money and were valued by what they did on the bottom line vs. whatever pipe dream they could come up with in a deck of PowerPoint slides. My favorite idea of all time remains a PowerPoint-to-Real-Product compiler.
Venture capitalists' portfolios went from returning stratospheric gains to losing money hand over fist. The rate of IPOs went from three a day to two a year. And along the way, Darwinism stepped in and did its thing. Real businesses had to remain real businesses. I truly thought the dumb stuff was behind us.
But alas, it seems we're destined to be dumb again. After a dearth of IPOs over the last five years, we've suddenly found ourselves back on the brink of stupidity. Riverbed Technology, a terrific company, successfully went public and has a market cap of $1.2 billion. The company--I really do love it--has small revenues and doesn't make money. I'm no Einstein, but $1.2 billion sure seems like a lot of dough. CommVault followed with another successful IPO. A good company in a tough market that hasn't made money raised a big pile of cash, most of which didn't go to the company but instead paid off investors. Double-Take Software is in registration and waiting to go out; it has made money for the last year, so I'm dying to see what kind of valuation it can pull off. Isilon Systems is also set to go.
There are a bunch of computer companies that have built annual revenues of $50 million to $100 million champing at the bit right now, which means there are a ton of venture capitalists sitting on boards telling CEOs to stop making money and start spending! Hire fast or you're fired. Spend like a drunken sailor on shore leave. Forget the fundamentals, there's gold in them thar hills!
Hey, VC/Wall St. guys, what if we encourage these companies to run real businesses and go public? How about valuing them based on what they achieve instead of what comes to you in a dream? I'm glad you're going to make money again, I really am. I'd be even happier if you could do it without being idiotic.
My latest CW rant. Sorry Carly.
I have lots to catch up on after spending ten days in a drugged out stupor post hernia and while you are there, lop off an extra inch or two, surgery. As soon as I catch up I'll blog on my visionquest experience in which I enjoyed some of the most bizzare hallucinations and utter clarity at the same time. The best part is that I was completely drug free, which is probably why it happened in the first place. More on that later.
This article is the follow up to the week before's - which generated a zillion responses. Seems to have ruffled some feathers.
Enjoy
Here's my CW article and here is the follow up because I couldn't stop writing - so this report on our site (it's free) is the complete ITIL/ITSM thing.
Basically, I wanted to have you understand what the stuff is, why it's important (for reasons other than what ITIL/ITSM folks say), and where it falls short. Let me know what you think.
Finally, I survived my first race track driving experience. Barely. In my "novice" class of 6, Aerosmith drummer Joey Kramer was one us. I'm a way better driver than him. When I asked him why he didn't bring his white Lambo Gallardo spider, he said "track soot". Track soot? Does he not realize there are probably lots and lots of groupies that would wash his car, or anything else for that matter, hoping to get it turned into an MTV video? Silly drummer.
It was one of the more terrifying, yet thrilling, experiences of my life. I did not destroy my car, which is nice.
My latest StorageMagazine rant on the worst reasons yet to have screwed up IT capabilities - because your boss's boss's boss is an idiot. If you can't see it due to passwords, yell at them. Since I own the content from now on I'll post the original article (unedited) so you don't have to go thru password hell.
Reyes is out on $2M bail, which is the equivilant of $1.87 to you and me. Federal judge denied his motion for dismissal. This has nothing to do with the article, but thought i'd mention it.
My thoughts on doing business with the storage vendors - my latest in ComputerWorld.
I've been writing a lot about virtulalization lately, and the process has been clarifying. I figured out a few things, such as the fact that none of the stuff we are talking about is new. Most of it is really old actually.
Perhaps the most interesting conclusion I've come too is that all virtualization (computer, anyhow) solutions we've ever come up with have followed identical financial paths. First, they attempt to solve capital expense related problems. Second, the attempt to solve operating expense related problems. And third, somewhat related to second, they attempt to keep people from jumping off buildings - or solving people scale problems.
A storage example would be Volume Manager. VM came about because of capital economics - i.e. when Seagate came out with a 4GB disk drive that was only 1.5X the cost of a 2GB disk drive, people bought them - lowering their capital cost per MB. The problem was that the operating system could only deal with a 2GB disk, so in reality what you added was a more expensive 2GB disk - until VM made it look like (2) 2GB disks. Phase two of storage virtualization was when we came out with RAID arrays - taking a ton of disks and putting them into a single box making it look like whatever we wanted. By doing that we got huge operating efficiencies, because managing a million individual disks is harder than managing one big box. Now we are at stage three, where we have a million big boxes, and people are going to jump of the building.
We are focused on keeping people from killing themselves, so most of the solutions to the Multi-box management problem are just moving the virtualization element to the next step in the process - looking at the boxes instead of the disks, etc. That will help, but it's not the end game, it just is like all the others, a band-aid.
Servers are now doing what storage did years ago. Servers cost too much and we have too many, so use virtual machines instead. Smart. Not new, but smart. Mainframes and big Unix boxes have been doing this since the dawn of time. Now we can do it on Windows and you'd think we've seen the second coming.
What we need is the 4Th motivation -which no one has figured out yet - and that is the ability to virtualize AND control the entire layer - be it storage, server, network, or whatever. That means much more than making one thing look like something else - it requires deep, intimate knowledge of all the underlying infrastructure so that while it masks the complexity from you and I, behind the scenes it controls it - the way an operating system does in your PC. Then you have real business value - that's when we stop doing tactical IT infrastructure stuff and all our efforts go into doing something with the data we have to drive down costs, make more money, or solve world peace - something a tad more meaningful than mastering an array GUI.
This is hard, of course, as recently pointed out in this blog by Greg Nawrocki who comments about another claim about virtualization and cold fusion, by upstart Crosswalk. I read the Crosswalk release, and heck it sounds awesome, but even if you nailed one layer, you have to be able to have understanding deep enough to coordinate between all the primary virtualization layers - be they server, storage, or network.
Howard Smiths blog, Forked Path talks about all sorts of stuff that shows the difficulty of the big picture interdepencies.
This is a good virtualization specific site that spends more time on server side stuff, but it's stuff you'll need to know.
A big part of the problem ultimately is that the different virtualization layers have much different players, and they are solving much different problems without much concern about the other. If you build a hammer, everything looks like a nail, right? 15 years ago you bought your system, application, network, and storage from the same guy, but no one does that anymore. They were forced to deal with each other, or perish. The virtualization guys - who's market is still too nascent, don't have to deal with each other yet. VMware is about a close as it comes since they are owned by EMC - but if you talk to an old time VMwarian, they don't know jack about storage, and certainly nothing about networks.
It's really hard to make things really easy.
Hu Yoshida, of HDS fame, spoke about a conference in his latest blog, and one of the topics was virtualization.
Hu says that users at a CIO panel commented that "the next step in virtualization is virtualized environments. Where you can swap out a compliance environment for instance and then bring it back later when it is needed".
Exactly. When we speak about virtualization in specific tactical terms, such as storage virtualization for migration purposes, or server virtualization for consolidation purposes, we lose the higher potential of the concept - which is really to do two things: First, it should abstract the user from the infrastructure on thier way to and from the data they care about, and second, behind that abstraction should be a living, breathing, morphable blob that can alter itself in order to best fullfil the requirments from the top of the stack (user) or the bottom (data).
I like the fact that people are talking about the V word more openly, and with less visible disdain - even if it is in terms that are still too simplistic - such as "improved utilization". Eventually people will come to grips with the fact that a fully integrated "Enterprise IT Virtualization" strategy will be the IT equivilant of the industrial revolution.
I wrote more on it in this weeks CW column.
"67% of all statistics are made up." Greg Duplessie
It does seem like that doesn't it. Remember when SRM was going to be a $11 Billion dollar business in 2008, or something?
Did you read Tony Asaro's latest? In his blog he talks about the end of SAN and NAS - and he's right. Do you think anyone really cares about SAN or NAS? If you aren't in the storage business, it's a good bet that you don't.
We are way too hung up on what we call things, instead of what they do - or better yet, what we need them to do. No one wants a Storage Area Network - they want a way to have a giant pile of disks be centralized from a management perspective that can act like a ton of direct attached storage mini-piles. The rationale is sound - if we pile them all together we can have them managed by less people and (in theory) we can get better utilization - i.e. better return on assets, out of one big pile vs. a zillion little piles. Do I really care if it's a fibre channel block storage network or an Ethernet based file system interface? Again, not unless you are in this business.
Tony's point, is that you shouldn't need to care - the storage should be smart enough to feed the application requests in the most efficient means possible, and you shouldn't have a blessed thing to do with it. The machine is smarter than you, at certain things. This should be one of them.
It wasn't long ago when part of the system admin's job was to assign physical memory to an application. You assigned page and swap space too. When is the last time you did that? Exactly - the machine is much better at making those decisions than you are. Your job is to figure out what application will lend the most value to whatever area of the business you are in - and then to establish the most effective means of delivering that application value to your consumer - the employee. Your value shouldn't be deciding if it's NAS or SAN any more than it was to figure out the optimal size page file for All-In-One.
In a lot of ways, the storage industry is the most back-assed of all. We have more high-tech stuff jammed into higher-tech packages that forgot the most important lesson of all - how to take the poor human out of the process. Imagine getting in your brand new Infiniti MX-35 - a high-tech car beyond compare. It has every conceivable gizmo, from a bluetooth cell phone connection that automatically routes calls thru the sound system, to a navigation system you can view in 3D. Now imagine that each time you got into it, it asked you to calculate exactly how much gas you would need to get home, based on today's price, once you manually figured out exactly how far you had to go. It would make the sexy new high-tech vehicle seem like a giant pain in the you know what. That's kind of what storage is.
We'll really be on to something when we stop asking if you need 4Gb/second or 2, and start asking "what are your objectives?" and then letting the machines figure out the best way to attain them.
In September 2004, ESG published a report entitled the Future of Network Based Storage Intelligence (you can find it at www.enterprisestrategygroup.com - it may cost you money to see it). Since then, a lot has happened - but one thing of note hasn't - the market.
Now, since we are always right (eventually), I'm trying to figure out what the deal is. We said that moving block storage services from the arrays and servers into the network would be one of the most important storage paradigm shifts since the advent of the block storage network itself. That statement is still true, so why isn't it commonplace yet?
The technical rationale for such a move hasn't changed - to consolidate and standardize disparate implementations of storage services that currently reside on various arrays and various servers to a centralized, standardized platform(s). Services such as volume management or data copy or data migration are handled differently by different operating environments and platforms. That means different people with different skills need to know different things. Different doesn't scale.
So if the motivation is valid, and the direction clear, then why aren't we there yet? Good question - there are several factions to blame.
First, the place this stuff should reside is on a switch. The switch isn't a "nice to have" but a "need to have". It's the junction box of all roads IT - everything traverses the switch, and as such, that is the place to to do deterministic things. The switch guys dramatically underestimated the engineering effort it would take to make their platforms capable of layering intelligence on their boxes. They figured that since they already did hard stuff - cracking packets at wire speed type stuff, that creating a replicate packet and shipping it to some other location couldn't be that tough. They thought "how hard can simple volume management be to run?" It turns out it was hard.
I'm happy to (finally) report that the smart switch guys appear ready to rock. Cisco and McData are finally ready, and Brocade not only seems ready - but (yes, this is me saying this) they seem to be leading. They already have their "data mobility" ARM software out running on their Rhapsody platform all over the place - and it appears to be working fantastically. Anyhow - all of them are going to go prime time with storage intelligence capabilities this year - 2006.
The second issue has been the software guys. The mainstream guys got all hot and bothered initially, and put a lot of effort into porting their stuff to the switch guys, but when the switch guys ran into the brick development walls, the ISV's put the brakes on. I don't blame them - they already have platforms to sell onto.
EMC and Incipient were the only two real software development efforts left standing in the last year or so. EMC had its own issues, as hardware companies turned software juggernauts will have, but finally is out with InVista and pushing it madly. From what I can see, the early production results are tepdily positive - people are rolling out InVista. Incipient is really the only true independent software player in this market (EMC stuff is designed to support EMC stuff, believe it or not). The good news is they have no other business to distract them, the bad news is until some of these switch guys get their act together and ship some boxes, Incipient has no place to sell its wares. Now that the box guys are finally in the game, I suspect we'll start seeing some real momentum.
What will be interesting now that smart switch fabrics are starting to happen in earnest, is who will be too slow to react? EMC by default will come out the top guy initially - they will do the usual thing - put a massive amount of effort and resources into building the market. We know IBM is a player already with SVC, and we think they are getting ready to do real battle via their Cisco relationship. They hate EMC, and will not sit idly by while InVista tries to steal their girl, so to speak. HP's strategy has been to draft off of IBM/Cisco (I presume with Incipient, but don't know for sure) which is smart - as long as they can react once the light turns green. McData and Brocade have all the footprint, but also have a problem with their customer base. The OEMs who sold all their gear have other agenda's, so finding the correct balance will be tricky. Regardless, McData and Brocade each should realize that the future depends on themselves - and they can't be forever kept as concubines if they want to survive and thrive. Sun should make a play - they are ones who said "the network is the computer" for crying out loud.
The winners will be the ones who get out first and stay put. This market will roll out exactly like the Veritas volume manager success story of days of yore. Veritas was brilliant giving away its volume manager to Sun - who shi |
