Al Shugart died this week - one of the brightest stars in the storage universe. Mike Cassidy's blog for the Mercury News does a nice job remembering him. Here's the url (link):
http://www.mercextra.com/blogs/cassidy/2006/12/13/al-shugart-dies-well-miss-the-genuine-article/
Shugart was replaced by Steve Luczo at Seagate in one of those cruel business turnabouts. Shugart brought Luczo in to help build a software business at Seagate through acquisition, which became a financial gambit in a larger plan of Luczo's. Its hard to argue with Luczo's vision and skill in taking Seagate private - probably not what Shugart envisioned, but a move that was every bit as brilliant as many of Shugart's had been over the years. Luczo was very impressive as Shugart's replacement but I've always felt he was a bit of a Brutus to Shugart's Julius Caesar.
Shugart's influence on the network storage industry was huge, but has been somewhat overlooked. IBM was getting ready to roll their Serial Storage Architecture (SSA) disk drives in the mid nineties, diminishing Seagate's influence in the enterprise disk drive business. Shugart and a team at Seagate went out to find a standards-based technology that they could use to compete with SSA and came across Fibre Channel. At the time FC was a floundering backbone network technology with a dim future due to the emergence of Gigabit Ethernet. The near miracle Seagate executed was building a consortium of companies that created a new, standard access method for Fibre Channel (FC-loop) that crushed IBM's SSA. As good as the technology feat was, the business creativity and leadership was way off the charts. Now I'm not much of an advocate for Fibre Channel anymore with iSCSI being so good, but its not obvious to me that SANs would have happened at all if not for FC-loop technology.
There is no question that Seagate's fortunes this decade would have been greatly diminished if not for their sizable high-margin FC disk business. IBM never really recovered from the loss and eventually sold its disk drive operations to Hitachi. Moreover, I don't think Luczo would have had the numbers to take Seagate private if the FC drive line hadn't been floating the company through an otherwise miserable cycle of chronic over-production.
Al Shugart was a true industry leader with a reach that was wide. Thanks, Al for your immense contributions to our industry and livelihoods. You really were the man.
December 16, 2006
Wii Redux - the thing runs on candle power
On the east coast for a business trip last week, I got a call from my despondent son. "Dad, the cat ate the wire from the sensor bar on my Wii."That really sucked, because the wire is super thin and although I told him I could fix it, I had my doubts about being able to solder the skinny little things together without screwing up the impedance and signal integrity in the wire. Considering how hard these things are to come by, I figured it would be almost impossible to get a replacement part any time soon.
When I started looking into the problem of a broken sensor bar, I found web sites discussing work arounds for the sensor bar - mostly from projection tv gamers who can't deal with the limited wire length (a shortcoming of the product - pun intended). The most interesting approach substituted a pair of candles for the sensor bar. This youtube video captured this setup working: http://www.youtube.com/watch?v=9ZxwBrCAtUE
I tried it myself and voila! it worked. My son was back in bliss, playing with a couple tea candles burning on the top of the TV like an altar decoration. Here's the deal. The sensor bar does not sense anything at all - instead it is a pair of infrared emitters (actually a pair of IR LED clusters). The Wii controller gets its location information from its position relative to those IR sources and then transmits that information to the console over its wireless connection. I made a couple sloppy solder joints to fix the wire and get power back to the sensor bar.
That's probably enough about gaming. I'll get back to storage now.
When I started looking into the problem of a broken sensor bar, I found web sites discussing work arounds for the sensor bar - mostly from projection tv gamers who can't deal with the limited wire length (a shortcoming of the product - pun intended). The most interesting approach substituted a pair of candles for the sensor bar. This youtube video captured this setup working: http://www.youtube.com/watch?v=9ZxwBrCAtUE
I tried it myself and voila! it worked. My son was back in bliss, playing with a couple tea candles burning on the top of the TV like an altar decoration. Here's the deal. The sensor bar does not sense anything at all - instead it is a pair of infrared emitters (actually a pair of IR LED clusters). The Wii controller gets its location information from its position relative to those IR sources and then transmits that information to the console over its wireless connection. I made a couple sloppy solder joints to fix the wire and get power back to the sensor bar.
That's probably enough about gaming. I'll get back to storage now.
November 29, 2006
Couch Potato Gets off Couch and Gets Sore Playing Wii
On black Friday, I got up at 4:45 AM and met my 12 year old son in the dark hallway of a house in Medford Oregon. 30 minutes later we were bleary-eyed and 5th in line at the local EB Games store waiting to spend my son's nest egg on the new Nintendo Wii (pronounced "we") game console. As I was standing there with other parents and their kids I couldn't help but think that we were all there to buy a machine that would be a divider between the generations. The irony of taking an unusual effort to buy one more anti-social escape mechanism was not lost on me. I was feeling a bit helpless that this new Wii, which my son saved his money for since last spring was destined to produce more couch-plated vegetating inactivity and increasing his risk of future blobdom.
Oops, wrong! I never thought I would be blogging favorably about a game console, but here it is. The Nintendo Wii is an extremely creative machine. While not as thrilling as an actual strap-in flight simulator, the Wii is amazingly stimulating and has the best interface ever imagined. Instead of exhausting the phalanges and tendons pushing buttons, the Wii is run by waving the controller around in the air. Not only that, but the darn thing seems to work best when you get the whole body behind the motions. Want to virtual bowl on the Wii? then practice your steps and take big, active swings with the controller. Want to score better at Tennis? then you ought to be up on your toes, paying attention to move to the next position to hit the ball on the screen. Want to box with a fictional opponent, then you better start ducking and weaving to avoid getting virtually knocked out.
The proof for me is the fairly widespread muscle soreness I have in my shoulder and back from waving the Wii's wand around. No kidding - this physical reaction is the real deal. The fact is, when you play the sports games on this thing, you get a little miniature workout. It will be very interesting to see what new action games are going to be like with this system. Its pretty clear to me that I'm going to have to start working on doing more with my left hand (and arm) in order to spread the fun and pain around a little.
So here's the final analysis: get one of these for somebody this holiday season if they 1) like gaming and 2) need more physical activity and 3) live close enough to you that you can clock some hours on it too. Its not an exercise machine, but it is a heck of a lot more fun than any exercise machine ever built and I suspect its owners won't stop using it because it's boring. It's definitely not boring. My son and I are now locked into virtual competition that is going to last for quite some time because I can actually compete with him. Rather than being a divider, this machine is giving us something to do together that's good for both of us. Wii = we. I never would have believed it.
Oops, wrong! I never thought I would be blogging favorably about a game console, but here it is. The Nintendo Wii is an extremely creative machine. While not as thrilling as an actual strap-in flight simulator, the Wii is amazingly stimulating and has the best interface ever imagined. Instead of exhausting the phalanges and tendons pushing buttons, the Wii is run by waving the controller around in the air. Not only that, but the darn thing seems to work best when you get the whole body behind the motions. Want to virtual bowl on the Wii? then practice your steps and take big, active swings with the controller. Want to score better at Tennis? then you ought to be up on your toes, paying attention to move to the next position to hit the ball on the screen. Want to box with a fictional opponent, then you better start ducking and weaving to avoid getting virtually knocked out.
The proof for me is the fairly widespread muscle soreness I have in my shoulder and back from waving the Wii's wand around. No kidding - this physical reaction is the real deal. The fact is, when you play the sports games on this thing, you get a little miniature workout. It will be very interesting to see what new action games are going to be like with this system. Its pretty clear to me that I'm going to have to start working on doing more with my left hand (and arm) in order to spread the fun and pain around a little.
So here's the final analysis: get one of these for somebody this holiday season if they 1) like gaming and 2) need more physical activity and 3) live close enough to you that you can clock some hours on it too. Its not an exercise machine, but it is a heck of a lot more fun than any exercise machine ever built and I suspect its owners won't stop using it because it's boring. It's definitely not boring. My son and I are now locked into virtual competition that is going to last for quite some time because I can actually compete with him. Rather than being a divider, this machine is giving us something to do together that's good for both of us. Wii = we. I never would have believed it.
November 12, 2006
Full speed ahead into V-Land
So now that I've had a chance to digest what I took in at VMWorld last week, I now see a much different future for IT organizations than I've seen before. The thing I like about it, is that this new vision didn't come from industry observers and pundits, but from customers who are taking sizable risks and accomplishing amazing things.
System virtualization really is the greatest thing since sliced bread and a lot of conference attendees told me so. They significantly simplify the management of their servers and applications by placing them into a consistent environment on top of VMWare ESX 3.0. Why would you want to manage lots of different physical systems, with all their quirks and personalities, when you can manage a much smaller set of servers running VMWare and providing consitent, vanilla virtual system configurations for all your applications? The answer is loud and clear - you wouldn't.
I talked to customers from Europe, Asia, South America, Canada and the US who are creating new, large server infrastructures on top of VMWare because it saves huge amounts of administrative time. I'm not joking when I say that we haven't seen a technology shift this radical since refrigeration became widespread. Once a system and application is placed in the VMWare environement it's functionality is sealed and preserved. The system hardware underneath the virtual system stack can change without impacting the performance or availability of the application. IT workers really like having control and not having to worry about new technologies screwing up the works. There is no way these people will ever go back to working the old-fashioned way.
A couple of times I heard VMWare customers at the conference say things like: "I don't care if they fire me, I'm not going to go back and make everything stupid again", or "I keep losing my best people because they can get better job offers after they learn how to do this - but its not as serious a problem as it used to be beause our virtual environment doesn't need as much attention."
One of our favorite customers, David Siles from Kane County, IL gave a presentation where he stated - among many other things - that one of the best reasons for putting applications on VMWare was to get disaster recovery protection for them that was too difficult or expensive to achieve in their native operating environments.
Wade O'Harrow, one of our SE's from the southeast was there and completely in his element as a VMWare animal. He kept describing our technology as "doing the same thing for storage as VMWare does for servers." I admit it took me awhile to figure it out, but I finally did get it. In an EqualLogic SAN, the individual storage systems are really secondary to the logical storage pools that customers create. Volumes can be moved among the underlying storage hardware resources - more or less independently of what those resources happen to be. You need disk drives for virtual storage just like you need processors for virtual servers, but each disk drive just gets less and less important in the scheme of things.
It seems weird to say that the great thing about our systems is that you don't have to depend on them individually to run your data center, but I'm getting used to this idea very rapidly as customers tell me that they don't depend on any of their servers either.
System virtualization really is the greatest thing since sliced bread and a lot of conference attendees told me so. They significantly simplify the management of their servers and applications by placing them into a consistent environment on top of VMWare ESX 3.0. Why would you want to manage lots of different physical systems, with all their quirks and personalities, when you can manage a much smaller set of servers running VMWare and providing consitent, vanilla virtual system configurations for all your applications? The answer is loud and clear - you wouldn't.
I talked to customers from Europe, Asia, South America, Canada and the US who are creating new, large server infrastructures on top of VMWare because it saves huge amounts of administrative time. I'm not joking when I say that we haven't seen a technology shift this radical since refrigeration became widespread. Once a system and application is placed in the VMWare environement it's functionality is sealed and preserved. The system hardware underneath the virtual system stack can change without impacting the performance or availability of the application. IT workers really like having control and not having to worry about new technologies screwing up the works. There is no way these people will ever go back to working the old-fashioned way.
A couple of times I heard VMWare customers at the conference say things like: "I don't care if they fire me, I'm not going to go back and make everything stupid again", or "I keep losing my best people because they can get better job offers after they learn how to do this - but its not as serious a problem as it used to be beause our virtual environment doesn't need as much attention."
One of our favorite customers, David Siles from Kane County, IL gave a presentation where he stated - among many other things - that one of the best reasons for putting applications on VMWare was to get disaster recovery protection for them that was too difficult or expensive to achieve in their native operating environments.
Wade O'Harrow, one of our SE's from the southeast was there and completely in his element as a VMWare animal. He kept describing our technology as "doing the same thing for storage as VMWare does for servers." I admit it took me awhile to figure it out, but I finally did get it. In an EqualLogic SAN, the individual storage systems are really secondary to the logical storage pools that customers create. Volumes can be moved among the underlying storage hardware resources - more or less independently of what those resources happen to be. You need disk drives for virtual storage just like you need processors for virtual servers, but each disk drive just gets less and less important in the scheme of things.
It seems weird to say that the great thing about our systems is that you don't have to depend on them individually to run your data center, but I'm getting used to this idea very rapidly as customers tell me that they don't depend on any of their servers either.
November 08, 2006
Where's the juice? VMWorld
I hate to admit it, but I've been a slacker when it comes to VMWare. I knew it was a big deal, but I never got that close to it before to do a deep dive and try to figure it out.
So after joing EqualLogic with my new mission to "Go forth and engage customers" I was directed to VMWorld in Los Angeles (Nov 6-9). I arrived in LA on a day when the temperature downtown was a record 97 degrees and I swear most of the heat was coming from the LA Convention Center where the show was.
VMWorld is a total buzz-o-rama, and I mean that in a very positive sense. This show has gathered many of the most wired people I have encountered in a long time. Now as a storage geek, I like storage shows and the other storage geeks that show up at storage shows, but I have to say that VMWorld is more amped up than any show I've been to for a long time - and almost everybody I met there that I knew said the same thing.
Our booth was gonzo. People were packed in. Some were customers coming buy to give a little love, some were Fibre Channel skeptics with "prove it" agendas, some were newbie customers with dig deeper questions and some were wired VMWare attendees just trying to drink in as much as they could. Tech talk was down and dirty into virtualized structures, app and data moves, drivers and disaster recovery scenarios. T shirts were flying off shelves and out of cabinets. It was head spinning. I haven't had a big adrenaline rush like this at a trade show - EVER. So props to the VMWare folks for making such an incredible product and building an amazing customer base.
If you are in LA and want to get together this week, let me know. 408.210.7931
mfarley@equallogic.com
So after joing EqualLogic with my new mission to "Go forth and engage customers" I was directed to VMWorld in Los Angeles (Nov 6-9). I arrived in LA on a day when the temperature downtown was a record 97 degrees and I swear most of the heat was coming from the LA Convention Center where the show was.
VMWorld is a total buzz-o-rama, and I mean that in a very positive sense. This show has gathered many of the most wired people I have encountered in a long time. Now as a storage geek, I like storage shows and the other storage geeks that show up at storage shows, but I have to say that VMWorld is more amped up than any show I've been to for a long time - and almost everybody I met there that I knew said the same thing.
Our booth was gonzo. People were packed in. Some were customers coming buy to give a little love, some were Fibre Channel skeptics with "prove it" agendas, some were newbie customers with dig deeper questions and some were wired VMWare attendees just trying to drink in as much as they could. Tech talk was down and dirty into virtualized structures, app and data moves, drivers and disaster recovery scenarios. T shirts were flying off shelves and out of cabinets. It was head spinning. I haven't had a big adrenaline rush like this at a trade show - EVER. So props to the VMWare folks for making such an incredible product and building an amazing customer base.
If you are in LA and want to get together this week, let me know. 408.210.7931
mfarley@equallogic.com
November 04, 2006
CAS: High cost and long term risk
Confusion and mystery are good things for vendors and bad things for consumers. Secret sauces are expensive. The storage industry has many examples of amazing profits borne on the back of opportunistic proprietary inventions.
The distinction between storage and data can be more difficult to ascertain than one might think. The concept of CAS (content addressable storage) is an example of how storage and data have been twisted together into a confused technology collage, creating an opportunity for vendors to print money.
At the end of the day, IT departments need to find ways to locate data. In the case of CAS, the data is placed in a special repository for historical archives and various algorithms are churned, creating resultant metadata that can be used to locate data at some later time. The belief is that preparing metadata today saves some amount of time and cost later.
Like many insurance or warranty products, CAS is paid for in advance. CAS customers pay top dollar for storage capacity that is likely going to be used very little. While there are probably some cases where customers use CAS products regularly, the question is: Couldn‚’t the same thing have been done at much lower cost? The answer to that question is obviously, yes. One could choose to use indexing, search or archive software tools, any of which could be more effective than what is integrated by the CAS vendor.
So, instead of placing archived data on expensive little-used CAS storage, it can be placed on low cost, little used storage where it is acted upon by any number of archiving, indexing or search technologies. The big advantage of separating software from hardware this way is that all the components of the archiving solution can be replaced if desired. If better software comes along, it can be easily incorporated. With a CAS system, customers are at the mercy of the hardware vendor to provide an update or upgrade.
The thing to keep in mind about archiving is that the function is expected to be needed far into the future, far past the lifespan of most technologies. The notion that any technology - storage hardware in particular - would have the ability to successfully manage the lifecycle of data archives is simply bizarre. The risk of having a data blackout in the future increases if succesful data access depends on the continued longevity of any particular vendor and any particular product sold by that vendor.
Archiving is going to be difficult work for many of us for many years, but it will be less difficult for those who don't paint themselves into corners with vendors looking for customers to milk with a string of expensive upgrades and services.
The distinction between storage and data can be more difficult to ascertain than one might think. The concept of CAS (content addressable storage) is an example of how storage and data have been twisted together into a confused technology collage, creating an opportunity for vendors to print money.
At the end of the day, IT departments need to find ways to locate data. In the case of CAS, the data is placed in a special repository for historical archives and various algorithms are churned, creating resultant metadata that can be used to locate data at some later time. The belief is that preparing metadata today saves some amount of time and cost later.
Like many insurance or warranty products, CAS is paid for in advance. CAS customers pay top dollar for storage capacity that is likely going to be used very little. While there are probably some cases where customers use CAS products regularly, the question is: Couldn‚’t the same thing have been done at much lower cost? The answer to that question is obviously, yes. One could choose to use indexing, search or archive software tools, any of which could be more effective than what is integrated by the CAS vendor.
So, instead of placing archived data on expensive little-used CAS storage, it can be placed on low cost, little used storage where it is acted upon by any number of archiving, indexing or search technologies. The big advantage of separating software from hardware this way is that all the components of the archiving solution can be replaced if desired. If better software comes along, it can be easily incorporated. With a CAS system, customers are at the mercy of the hardware vendor to provide an update or upgrade.
The thing to keep in mind about archiving is that the function is expected to be needed far into the future, far past the lifespan of most technologies. The notion that any technology - storage hardware in particular - would have the ability to successfully manage the lifecycle of data archives is simply bizarre. The risk of having a data blackout in the future increases if succesful data access depends on the continued longevity of any particular vendor and any particular product sold by that vendor.
Archiving is going to be difficult work for many of us for many years, but it will be less difficult for those who don't paint themselves into corners with vendors looking for customers to milk with a string of expensive upgrades and services.
October 31, 2006
Storage Customer Advocacy and My Role @ EqualLogic
So, I've done the unthinkable - I've decided to go to work for a company again. This time its EqualLogic, the storage manufacturer from Nashua NH that is best known as an iSCSI company, but their (our) technology goes too far beyond iSCSI to be limited by the assumptions that usually go along with iSCSI. Enough on that topic for now, there will probably be lots more on this topic in the months to come.
People have asked, why would a guy who has proclaimed to be unmanageable and unable to work for anybody else go back to work again in a corporate environment shlepping products? The answer is easy; as an analyst/free lance writer, the only way I was able to sustain an income was to shlep products for clients. I should mention that it is possible to make some money writing books, but its never as much as people think it is and the amount of work is extreme. Its not exactly a great business model. Writing for magazines and e-zines doesn't pay either. In the final analysis, I was destined to be a product shlepper.
In the course of my work, there are some products that I really got excited about such as EqualLogic's PS Series and IBM's SAN Volume Controller or Prostor's RDX, but there were a LOT of other products that couldn't fire my imagination to the same degree. Over time, I discovered I wasn't very good at shlepping mediocre products or companies - there was usually something far more interesting in my sphere such as a hot dog stand, video game or a nap.
So I was approached to go to work by somebody that I probably shouldn't mention - but it was real flattering and I was real interested and all that and I liked their ideas and their business and everything a whole lot, but I had made a promise to Don Bulens (CEO at EqualLogic) that if I ever considered going to work again, I would give him a shot at hiring me. So that's how it went down; somebody else got me tipped into thinking about going to work and Don created a position for me that I couldn't really refuse - customer advocate guy.
So now I get to go talk to customers and figure out what they want, what they don't want, talk to them about how they work, what interests them and catches their attention - what they think they will be doing 2, 3, 5 years from now and all the sorts of things people expected me to know as an analyst - but never had the ability to do. I can't believe that somebody is actually going to pay me to do this. Of course I can't believe that more companies don't hire more people to do this sort of thing.
So call or write if you get the urge, we can talk about all kinds of things: storage, business continuity, e-discovery, subsystem architectures, virtualization, the worst world series in history, cat behavior, remote copy and replication, favorite movies, dual parity RAID, whatever. I'm sure we can help each other and have a little fun along the way.
My contact info:
mfarley@equallogic.com
marc@buildingstorage.com
408-210-7931 (cell)
People have asked, why would a guy who has proclaimed to be unmanageable and unable to work for anybody else go back to work again in a corporate environment shlepping products? The answer is easy; as an analyst/free lance writer, the only way I was able to sustain an income was to shlep products for clients. I should mention that it is possible to make some money writing books, but its never as much as people think it is and the amount of work is extreme. Its not exactly a great business model. Writing for magazines and e-zines doesn't pay either. In the final analysis, I was destined to be a product shlepper.
In the course of my work, there are some products that I really got excited about such as EqualLogic's PS Series and IBM's SAN Volume Controller or Prostor's RDX, but there were a LOT of other products that couldn't fire my imagination to the same degree. Over time, I discovered I wasn't very good at shlepping mediocre products or companies - there was usually something far more interesting in my sphere such as a hot dog stand, video game or a nap.
So I was approached to go to work by somebody that I probably shouldn't mention - but it was real flattering and I was real interested and all that and I liked their ideas and their business and everything a whole lot, but I had made a promise to Don Bulens (CEO at EqualLogic) that if I ever considered going to work again, I would give him a shot at hiring me. So that's how it went down; somebody else got me tipped into thinking about going to work and Don created a position for me that I couldn't really refuse - customer advocate guy.
So now I get to go talk to customers and figure out what they want, what they don't want, talk to them about how they work, what interests them and catches their attention - what they think they will be doing 2, 3, 5 years from now and all the sorts of things people expected me to know as an analyst - but never had the ability to do. I can't believe that somebody is actually going to pay me to do this. Of course I can't believe that more companies don't hire more people to do this sort of thing.
So call or write if you get the urge, we can talk about all kinds of things: storage, business continuity, e-discovery, subsystem architectures, virtualization, the worst world series in history, cat behavior, remote copy and replication, favorite movies, dual parity RAID, whatever. I'm sure we can help each other and have a little fun along the way.
My contact info:
mfarley@equallogic.com
marc@buildingstorage.com
408-210-7931 (cell)
October 10, 2006
Google, You Tube and Community
As much as I hate to admit it, not long ago I didn’t even know what You Tube was. A friend of mine in the media business pointed me in their direction last Spring, telling me that this was a company that had the potential to shake the media world off it’s hinges. You Tube’s platform for personal expression is immediate and engaging. Based on the numbers and my own personal experience, the You Tube vortex appears to be very strong. In a year’s time, You Tube went from nothing to becoming the Internet’s reincarnation of Vaudeville, where both audience and players funnel into an electronic mosh pit of video surfing. If you go to look at one 30 second video can you keep from looking at another, and another and another and …….?
But more than just a place to exchange videos, You Tube lets everybody be a critic by giving voice to their opinions. Most of us at one time or another have probably thought we could do a better job than the seemingly humorless or incompetent movie critics in the world. Like most pundits (including technology analysts), there are a small percentage of movie critics who seem to know what they are talking about while the rest just seem to like the sound of their own voices. You Tube gives each of us the power to appreciate things on our own and provides us with a vehicle to critique. In the end, this is probably just as important as the ability to produce and publish videos. The kinds of things people communicate in You Tube are little intangible personal elements of their lives.
And that is what Google sees in the deal. They might be the most heavily used computer-utility in the world, but they haven’t had a way to compete with e-Bay for Internet-user mind share (Google users tend to pass through, whereas e-Bay users tend to spend a lot of time and shop). The thing that makes e-Bay work so well is its feedback system that establishes trust between total strangers. Google wanted to find a similar way for their users to engage each other and they believe they have found the intangible recipe with You Tube. There is no doubt that You Tube can get under your skin. The question is, would you buy a used car from somebody on Google partly because they like the same kind of videos that you do?
The message from Google here is clear: the most important thing in the world for a technology company is to build an excited community. If a company can do that, it can have irresistible market pull. This business philosophy applies broadly beyond Google, e-Bay and the websphere. Providing superior products and technical support are key elements to building an excited community of customers that trusts a vendor and its products and services. People tend to look to analysts for opinions on vendors and products. The analysts are pretty good at understanding the strengths and weaknesses of various vendors, but they seldom know much about products, how they work and how they are managed. If you want to know how a product works or how a vendor’s technical support is, ask a customer with experience. The best companies have the most excited and enthusiastic customers.
But more than just a place to exchange videos, You Tube lets everybody be a critic by giving voice to their opinions. Most of us at one time or another have probably thought we could do a better job than the seemingly humorless or incompetent movie critics in the world. Like most pundits (including technology analysts), there are a small percentage of movie critics who seem to know what they are talking about while the rest just seem to like the sound of their own voices. You Tube gives each of us the power to appreciate things on our own and provides us with a vehicle to critique. In the end, this is probably just as important as the ability to produce and publish videos. The kinds of things people communicate in You Tube are little intangible personal elements of their lives.
And that is what Google sees in the deal. They might be the most heavily used computer-utility in the world, but they haven’t had a way to compete with e-Bay for Internet-user mind share (Google users tend to pass through, whereas e-Bay users tend to spend a lot of time and shop). The thing that makes e-Bay work so well is its feedback system that establishes trust between total strangers. Google wanted to find a similar way for their users to engage each other and they believe they have found the intangible recipe with You Tube. There is no doubt that You Tube can get under your skin. The question is, would you buy a used car from somebody on Google partly because they like the same kind of videos that you do?
The message from Google here is clear: the most important thing in the world for a technology company is to build an excited community. If a company can do that, it can have irresistible market pull. This business philosophy applies broadly beyond Google, e-Bay and the websphere. Providing superior products and technical support are key elements to building an excited community of customers that trusts a vendor and its products and services. People tend to look to analysts for opinions on vendors and products. The analysts are pretty good at understanding the strengths and weaknesses of various vendors, but they seldom know much about products, how they work and how they are managed. If you want to know how a product works or how a vendor’s technical support is, ask a customer with experience. The best companies have the most excited and enthusiastic customers.
June 01, 2006
Sun Layoffs and Horse Flogging
SUN announced that it is going to layoff 5000 employees this summer. This of course is no great surprise. If you become an ex-SUN employee in the coming months you will need to decide what to do with any stock you might have. My advice is to sell, and invest elsewhere. It's possible that you might regret it, but it's equally possible that it might be a good move.
It's not Schwartz's fault that he inherited a mess. Perhaps he is figuring out how to unravel it, but perhaps he is just trying to squeeze numbers quickly and buy himself some time. I believe SUN is an early-stage turn-around project and saying that, the odds are heavily stacked against an insider being able to do everything that is needed. Just as IBM needed Lou Gerstner, SUN needed somebody who understands business first and technology - not much. Comparing SUN to IBM in 1993 is not so damning. There is a great deal that SUN can do, if it decides to go about it in a business-like fashion.
I am an OpenOffice user and my machine runs Java like every other machine - but I didn't pay anything to get it and it hasn't inspired me to buy any SUN products. People will spend money on SUN products and services if they think those products and services have value. SUN needs to drop its popularity contest software development efforts and leverage its heritage as a data center computing company (like IBM), but it needs to make decisions without the emotional attachments of historical precedence.
A month ago I wrote my thoughts on what SUN needed to do in storage. It was the same idea - that SUN needed an outsider to take a fresh objective look at the business. Obviously, I don't have much influence there because Schwartz hired company insider David Yen for the job. Yen has an impressive resume and knows a great deal about high performance computing and probably all other aspects of computing/networking - even storage. But I don't think he will be able to build SUN storage business up from its current shambles and make a go of it. He does not appear to have any history doing that sort of thing.
So, despite it's appearances, I can't help asking if Yen's appointment isn't actually a form of demotion. Old school whiz kids like David Yen never wanted to work on storage with the knuckle draggers. What's changed? More importantly, can somebody who probably delegated storage stuff to people who delegated it to somebody else sustain the energy and emotion needed to build a business in an intense industry like storage?
I get the feeling that if I met David Yen, I would probably like him and would undoubtedly respect his knowledge. But SUN just has everything wrong when it comes to storage. If SUN is an early stage turn around, then its storage business is a later stage massive pruning project (and turn around). Yen probably doesn't know that's the job and when he finds out in the coming months, is he going to embrace the challenge?
I see Schwartz, an insider, unable to do the radical things needed to remake SUN. I don't see him extending beyond his tech-guy comfort zone to reach into the icy blue cold of business logic to make his decisions. This layoff announcement didn't prove anything.
It's not Schwartz's fault that he inherited a mess. Perhaps he is figuring out how to unravel it, but perhaps he is just trying to squeeze numbers quickly and buy himself some time. I believe SUN is an early-stage turn-around project and saying that, the odds are heavily stacked against an insider being able to do everything that is needed. Just as IBM needed Lou Gerstner, SUN needed somebody who understands business first and technology - not much. Comparing SUN to IBM in 1993 is not so damning. There is a great deal that SUN can do, if it decides to go about it in a business-like fashion.
I am an OpenOffice user and my machine runs Java like every other machine - but I didn't pay anything to get it and it hasn't inspired me to buy any SUN products. People will spend money on SUN products and services if they think those products and services have value. SUN needs to drop its popularity contest software development efforts and leverage its heritage as a data center computing company (like IBM), but it needs to make decisions without the emotional attachments of historical precedence.
A month ago I wrote my thoughts on what SUN needed to do in storage. It was the same idea - that SUN needed an outsider to take a fresh objective look at the business. Obviously, I don't have much influence there because Schwartz hired company insider David Yen for the job. Yen has an impressive resume and knows a great deal about high performance computing and probably all other aspects of computing/networking - even storage. But I don't think he will be able to build SUN storage business up from its current shambles and make a go of it. He does not appear to have any history doing that sort of thing.
So, despite it's appearances, I can't help asking if Yen's appointment isn't actually a form of demotion. Old school whiz kids like David Yen never wanted to work on storage with the knuckle draggers. What's changed? More importantly, can somebody who probably delegated storage stuff to people who delegated it to somebody else sustain the energy and emotion needed to build a business in an intense industry like storage?
I get the feeling that if I met David Yen, I would probably like him and would undoubtedly respect his knowledge. But SUN just has everything wrong when it comes to storage. If SUN is an early stage turn around, then its storage business is a later stage massive pruning project (and turn around). Yen probably doesn't know that's the job and when he finds out in the coming months, is he going to embrace the challenge?
I see Schwartz, an insider, unable to do the radical things needed to remake SUN. I don't see him extending beyond his tech-guy comfort zone to reach into the icy blue cold of business logic to make his decisions. This layoff announcement didn't prove anything.
May 31, 2006
My Product of the Year
A couple months ago at Storage Networking World I saw a new storage product with great potential, but I couldn't talk about it because it was not announced yet.
Prostor, a startup in Boulder with a team of industry veterans has developed the ultimate tape replacement product - a removable disk cartridge that operates like a tape. In other words, it is a virtual tape cartridge that looks and feels just like a real tape cartridge, but it has a 2.5" disk drive inside. That means instead of having a complex and costly tape transport to deal with, the Prostor RDX (removable disk -xfactor?) system uses the highest volume/lowest cost disk interface and connectors in the universe. That alone should make the RDX the most desired backup technology ever created.
Of course, the big thing that most people will like about RDX is the lack of tape. The RDX does not have tape that flakes apart over time, tape that can't stand temperature and humidity changes, tape that can't stand magnetic coercion, tape that sometimes gets kinked up in the transport and needs to be removed with a tweezers, tape that needs to be retensioned every year when sitting on the shelf storing archives, tape that breaks off the end of the spool when you least expect it and tape that needs to be discarded because its header has been updated a couple thousand times already (this happens much faster than its sounds). Yes, tape has lots of problems and most of us can't stand any of one of them, much less the whole lot of them together waiting to ruin our days, jobs, careers, etc.
In the last few years, disk to disk (D2D) backup has become popular because tape is such a nuisance. The problem is, D2D backup with larger subsystems tends to be a bit pricey and the whole backup operations needs to be altered to incorporate an intermediate disk layer. Do you write to disk and then to tape? If so, can you recover directly from tape or do you need to recover to disk first? How is media managed when you write to tape from disk? Do these tapes use the same naming conventions and does the backup database know about these tapes? As you go further down the path with D2D and D2D2T and all variations, there are best practices and disciplines to develop that are thornier to deal with than it looks at first glance. Its worth doing, but its more work. Backup and recovery is always work.
So along comes Prostor with RDX and the next thing you know, there is no intermediate disk stage. You write to their disk cartridges directly during backup (yes - you need an RDX "drive" or robotic controller mechanism too) and naming conventions are taken care of without figuring out complicated schemes. There is no second backup-to-tape operation and there is no lag time required before you can take data off site to comply with disaster recovery edicts.
For people who have struggled with backups for years, this is the sort of thing you have been dreaming about. Run don't walk to Prostor's web site: http://www.prostorsystems.com/ Then start trying to figure out how you are going to get your hands on the stuff so you can start making your lives a whole lot easier.
There will be bumps along the way I'm sure like there are with any new technology, but the bumps this time likely to be a lot smaller and fewer than we've seen with tape over many decades of refining and suffering.
-Marc Farley
Prostor, a startup in Boulder with a team of industry veterans has developed the ultimate tape replacement product - a removable disk cartridge that operates like a tape. In other words, it is a virtual tape cartridge that looks and feels just like a real tape cartridge, but it has a 2.5" disk drive inside. That means instead of having a complex and costly tape transport to deal with, the Prostor RDX (removable disk -xfactor?) system uses the highest volume/lowest cost disk interface and connectors in the universe. That alone should make the RDX the most desired backup technology ever created.
Of course, the big thing that most people will like about RDX is the lack of tape. The RDX does not have tape that flakes apart over time, tape that can't stand temperature and humidity changes, tape that can't stand magnetic coercion, tape that sometimes gets kinked up in the transport and needs to be removed with a tweezers, tape that needs to be retensioned every year when sitting on the shelf storing archives, tape that breaks off the end of the spool when you least expect it and tape that needs to be discarded because its header has been updated a couple thousand times already (this happens much faster than its sounds). Yes, tape has lots of problems and most of us can't stand any of one of them, much less the whole lot of them together waiting to ruin our days, jobs, careers, etc.
In the last few years, disk to disk (D2D) backup has become popular because tape is such a nuisance. The problem is, D2D backup with larger subsystems tends to be a bit pricey and the whole backup operations needs to be altered to incorporate an intermediate disk layer. Do you write to disk and then to tape? If so, can you recover directly from tape or do you need to recover to disk first? How is media managed when you write to tape from disk? Do these tapes use the same naming conventions and does the backup database know about these tapes? As you go further down the path with D2D and D2D2T and all variations, there are best practices and disciplines to develop that are thornier to deal with than it looks at first glance. Its worth doing, but its more work. Backup and recovery is always work.
So along comes Prostor with RDX and the next thing you know, there is no intermediate disk stage. You write to their disk cartridges directly during backup (yes - you need an RDX "drive" or robotic controller mechanism too) and naming conventions are taken care of without figuring out complicated schemes. There is no second backup-to-tape operation and there is no lag time required before you can take data off site to comply with disaster recovery edicts.
For people who have struggled with backups for years, this is the sort of thing you have been dreaming about. Run don't walk to Prostor's web site: http://www.prostorsystems.com/ Then start trying to figure out how you are going to get your hands on the stuff so you can start making your lives a whole lot easier.
There will be bumps along the way I'm sure like there are with any new technology, but the bumps this time likely to be a lot smaller and fewer than we've seen with tape over many decades of refining and suffering.
-Marc Farley
April 25, 2006
Sun Storage: Will the Oxymoron Continue?
If you believe that the storage industry benefits from strong management with the desire to compete, Scott McNealy's decision to step aside at Sun last week was long overdue. Despite any sales claims Sun might offer, the storage business at Sun has been the laughingstock of the industry for many years. With the changing of the guard at Sun, one might think that Sun would try to create new growth opportunities for itself by leveraging its enormous install base and selling storage products and services to their customers. One might think that Sun would look to HP for a few clues on restarting a storage business. It's possible, but I wouldn't bet on it.
Sun's history with storage has been a stumbling string of arrogance, ignorance and incompetence. They were too arrogant to realize that people didn't like installing their relatively expensive and complicated systems when all they wanted was a simple file server. Instead of recognizing a clear opportunity and capitalizing on it, they watched with disdain for over a decade as Network Appliance grew up and became a large company with products that Sun was too proud to understand and emulate.
There is little question that a company with the talent to create the excellent Solaris Operating System could have also developed strong volume management and file system software. For reasons I will never understand, they handed the golden goose to Veritas, which built the dominant storage and data management software company as a SUN OEM development partner. This would have been important intellectual property for SUN - and could have helped them create important storage networking products - but they mysteriously abdicated the opportunity.
Then there was the great SAN misfire. Sun was the first big name systems company with Fibre Channel products, but Sun initially saw FC as a long distance storage connectivity option and not as a new kind of network. They later changed their perspective, but failed to gauge the market requirements correctly, missing the target with underwhelming products. In trying too establish a unique identity in an industry with compatibility problems, Sun's SAN solutions became technology investment traps. This was certainly ironic for the leading company of the Open Systems movement.
Now, let's fast forward to approximately one year ago when Sun decided to buy Storage Tek (STK) for $4.4 Billion. At the time I commented that this was money that Sun probably should have paid out as dividends to stockholders. STK did not have any special new technology, nor did they have an orthogonal install base to create new customers. STK did have cash and they had what appeared to be a sizable services business in storage that Sun could presumably use to create a larger storage services business. The problem is, a heavy component of STK's service business was generated from maintenance on their aging mainframe tape equipment, which at best could only be useful for tactical business purposes. $4.4 Billion is a lot to spend for something without strategic value.
Apparently Jonathan Schwarz, Sun's new CEO pushed for the STK deal. Its not clear what he was looking for - perhaps some notoriety as a deal maker. It's not so difficult to understand how working in the outspoken McNealy's shadow could be challenging - and you can't blame an ambitious executive for wanting to establish his own authority and vision. It's just that the STK deal was so underwhelming, it doesn't inspire much confidence in Schwarz's abilities to navigate the storage marketplace.
Many Sun observers are looking to see if Schwarz has the courage and resolve to cut headcount at Sun. For me, a storage person, I'm looking to see if Schwarz will restructure the storage business. Sun badly needs a fresh perspective in storage with new leadership that has the authority to do what is needed. I'm not likely to be much of a believer until Sun brings in help from outside the company who have a Sun-less outlook on the storage business and can make objective assessments of the resources at hand. The pruning within Sun storage needs to be very aggressive and determining who stays and who goes would be a very difficult job. It starts with finding a leader who can cut through the internal politics and history and focus completely on market opportunities. But before that can happen, Schwarz needs to realize the necessity of cleaning the storage house – and before that can happen, Schwarz needs to commit to making storage a key part of his business strategy. As a Sun insider who learned from watching Scott McNealy, that's not likely to happen anytime soon.
-Marc Farley
Sun's history with storage has been a stumbling string of arrogance, ignorance and incompetence. They were too arrogant to realize that people didn't like installing their relatively expensive and complicated systems when all they wanted was a simple file server. Instead of recognizing a clear opportunity and capitalizing on it, they watched with disdain for over a decade as Network Appliance grew up and became a large company with products that Sun was too proud to understand and emulate.
There is little question that a company with the talent to create the excellent Solaris Operating System could have also developed strong volume management and file system software. For reasons I will never understand, they handed the golden goose to Veritas, which built the dominant storage and data management software company as a SUN OEM development partner. This would have been important intellectual property for SUN - and could have helped them create important storage networking products - but they mysteriously abdicated the opportunity.
Then there was the great SAN misfire. Sun was the first big name systems company with Fibre Channel products, but Sun initially saw FC as a long distance storage connectivity option and not as a new kind of network. They later changed their perspective, but failed to gauge the market requirements correctly, missing the target with underwhelming products. In trying too establish a unique identity in an industry with compatibility problems, Sun's SAN solutions became technology investment traps. This was certainly ironic for the leading company of the Open Systems movement.
Now, let's fast forward to approximately one year ago when Sun decided to buy Storage Tek (STK) for $4.4 Billion. At the time I commented that this was money that Sun probably should have paid out as dividends to stockholders. STK did not have any special new technology, nor did they have an orthogonal install base to create new customers. STK did have cash and they had what appeared to be a sizable services business in storage that Sun could presumably use to create a larger storage services business. The problem is, a heavy component of STK's service business was generated from maintenance on their aging mainframe tape equipment, which at best could only be useful for tactical business purposes. $4.4 Billion is a lot to spend for something without strategic value.
Apparently Jonathan Schwarz, Sun's new CEO pushed for the STK deal. Its not clear what he was looking for - perhaps some notoriety as a deal maker. It's not so difficult to understand how working in the outspoken McNealy's shadow could be challenging - and you can't blame an ambitious executive for wanting to establish his own authority and vision. It's just that the STK deal was so underwhelming, it doesn't inspire much confidence in Schwarz's abilities to navigate the storage marketplace.
Many Sun observers are looking to see if Schwarz has the courage and resolve to cut headcount at Sun. For me, a storage person, I'm looking to see if Schwarz will restructure the storage business. Sun badly needs a fresh perspective in storage with new leadership that has the authority to do what is needed. I'm not likely to be much of a believer until Sun brings in help from outside the company who have a Sun-less outlook on the storage business and can make objective assessments of the resources at hand. The pruning within Sun storage needs to be very aggressive and determining who stays and who goes would be a very difficult job. It starts with finding a leader who can cut through the internal politics and history and focus completely on market opportunities. But before that can happen, Schwarz needs to realize the necessity of cleaning the storage house – and before that can happen, Schwarz needs to commit to making storage a key part of his business strategy. As a Sun insider who learned from watching Scott McNealy, that's not likely to happen anytime soon.
-Marc Farley
Subscribe to:
Posts (Atom)