Silicon bites the dust

When a certain product helps you meet the future Prime Minister of a country, it is not surprising that you remember the product fondly. That is why I turned a bit nostalgic when I read that Silicon Graphics’s assets have all been bought by Rackable for just $25million dollars. Afterall, Silicon Graphics Iris workstation took me to many places in India and no wonder that I feel like writing about it now. So excuse my self indulgence and read on.

It was in late 1980s when OMC computers and Wipro wanted to be the guys who sold Sun workstations in India. HCL was selling Appollo workstations. I had just joined OMC Computers and people told me that we lost out Sun to Wipro. It was good deal for Wipro because they sold a lot of Sun workstations. In order to compete in the workstation market, OMC tied up with Silicon Graphics, then one of the leading graphic workstations in the market. As we found out the hard way, given its price and positioning, it was a major challenge to sell it in India and in many cases Wipro beat us with the Sun workstations, which is what many enterprises wanted. I am not complaining since I got a lot of good experience trying to sell and support the Silicon Graphics workstations. For one, I got to meet lot of interesting people starting from academicians, film producers and all the way upto the future Prime Minister. Along the way, I had lot of interesting experiences as well.

It was clear to us after our efforts to fight Sun, that we cannot position Silicon Graphics workstations as general purpose workstations. (Wish the guys who made the deal knew it earlier.) So our strategy became more product or solution oriented. One of the segments we attacked for molecular biology since Center for Molecular Biology (CCMB) was located in Hyderabad, our headquarters. And next to it was RR labs (now IICHT). So we met lot of professors here and gave them demo on a certain product, whose name I cannot recall, to all of them. This was to enable some sort of 3D modelling of the proteins. It was then that I bought a book on Bio Chemistry and learnt that there were some 20 odd Amino acids and that all proteins were formed out a certain pattern of these amino acids. Looks like the scientists know the sequence of amino acids in the proteins but do not know about the actual physical structure of the protein. This software was supposed to help them solve this problem. Encouraged by the good words the profs had for this software, we decided to an all India roadshow and asked the company which had this product to send us an expert. They agreed and sent us the ‘expert’, who was actually a student and was doing a summer internship with them and was on a vacation to India!!  She was now called in to face a lot of academicians, who have been working in this field for ages. The encounter had the girl almost in tears. So I had to step in and control the situation. Like how we do always, I took many of the discussions ‘offline’, promised that we will send them more details later and did all the things we do when we don’t know an answer and don’t want to admit it. I remember my manager commending me later about how I was able to save the situation but it was not a situation I want to get in often. Lessons were learnt  in this encounter and  when we went to the next venue, were were prepared and things went off smoothly.

Along with the Molecular Biology software, we were also trying to sell some 3D modelling and animation software. I think it was called Alias. My colleague, Surya and I decided that we should try and sell this software to Annapoorna Studios, which was the biggest studio those days in Hyderabad. We ended up meeting Akkineni Venkat, the brother of the famous Telugu film hero, Akkineni Nagarjuna. He had visited our premises along with Ramprasad, the proprietor of Walden Book Store, and we gave them a demo on 3D modelling and showed some tricks like turning positive into negative etc. What we didn’t understand at that time was that whatever we had was not enough and to get a good graphics based stuff for films required lot more that just a Silicon Graphics machine and some software. We never made any deals with Annapoorna studios but the encounter did provide me with some memorable moments. One such moment was when Akkineni Venkat, who was watching a demo, stepped out to make a call. He called Nagarjuna’s house and apparently it was Nagarjuna’s wife and the beautiful actress Amala, who was on the line. I still recall Surya pulling my hand excitedly and saying, “He is talking to Amala!! He is talking to Amala!!”.

Then came the very brief encounter with this person. We were told by our manager that we need to put up the Silicon Graphics workstation for a demo on a Sunday. Obviously we were pissed. First, it was on a Sunday. Second, it was not a computer exhibition. It was an exhibition of all the companies to whom State Bank of India had provided funding. OMC Computers happened to be one of them. SBI wanted to show the good work it was doing to the then Finance Minister, who went by the name of Manmohan Singh. There we were, on a Sunday, waiting for the FM to arrive. He came in and started doing the rounds. What was impressive was that he took his time at each stall and spoke to people to get their direct feedback. He came to out stall and I was holding fort there. Those were the times of high export duties and Silicon Graphics was a costly system due to these duties. The FM was in our stall and we showed a small demo. After having a look he asked, “How much does this system cost?”. I replied, “Depends on your policies, sir.” This brought a small smile on his lips. He moved on after asking if people thought the cost was too high. What was striking was his utter simplicity and a real urge to understand the problems. No wonder he is considered to be one of our best Finance Minister ever.

Silicon Graphics took me to various corners of India. I experienced the searing heat of Jamshedpur, the freezing cold of Delhi, the sweating in Calcutta, the pleasantness of Bangalore. It is very sad to see Silicon Graphics bite the dust. I have not worked on it since I quit OMC Computers in 1995 but still, it seems like I lost a good friend.

A conference and a few cribs

“Why have a blog if you cannot rant once in a while?” That seems to be the motto of all bloggers. It is almost as if you will not be considered a blogger if you don’t give your piece of mind about what is wrong  with this world once in a while. I rant and rave now, so that I too will be considered a serious blogger!!

I attended a conference last week and I am going to talk about a few irritants. While I talk specific to this conference, it is probably true about most of the conferences that I attend. The first thing that bugs you are these 50,000 feet views. I can understand some CXO talking about what direction the industry is heading and and all that but it becomes a pain when every marketing person starts giving his or her view as the opening address. There is generally very less substance in the talk and everyone is waiting for the tea break  so that they can have some cookies and a cup of tea. The speaker of the keynote address wanted to be very interactive and asked the audience, “What is the most important asset of any company?” Since he belonged to the storage industry he was hoping that the audience will respond with “Data is the most important asset.” Instead the audience in one voice responded, “Employees.” That is the last thing any manager wants to hear in these troubled times, when knives are drawn and you don’t know whose neck you have to chop off next. So he manfully continued. “Yes. People are an important asset. What else?” Some replied, “Data” and put him out of misery. Then started the 50,000 ft view. My request to all the keynote speakers is, unless you are a CXO, don’t make your talk so generic that it is of no use to anyone.

The speaker also made another interesting observation. “You may have heard of ILM. You don’t hear about it much nowadays. Why?” The answer is that the customers have realized that they have to do the bulk of work in ILM and the vendors would just come in with their regular products labeled as ‘ILM ready’. Not according to the speaker. “Nowadays many of the array can do automatic tiering. This is exactly what ILM is all about. So your ILM has been pushed into the array and that is why no one talks about ILM.” Wish life were so simple!!

Another request to all speakers. Please do not assume that the audience is so dumb that you need to simplify everything to ridiculous levels. One of the speakers was explaining about Data Deduplication. In order to ensure the audience understood this technology, he said something like this. “Take the case of water molecule. It contains Hydrogen and Oxygen. So you need not save lot of water molecules. All you need is Hydrogen and Oxygen and we can make whatever quantity of water that we need.” Honest.  I am not making this up. Another speaker emphasized the same fact differently. “You need not save details of all materials. After all, every material is made out of some 100 and odd elements. If you have the basic elements stored, we can make whatever substance we want. That is what Data Dedup is all about.” This would lead the audience to conclude that since every computer stores data in binary form, 1s and 0s, all that which needs to be stored for every file is one 1 and one 0. Exactly two bits!! You can then recreate any data that you want from these elements!! I don’t think it will do any harm in putting up a few bit patterns or talking about hashing instead of oversimplifying the concept so much that it can easily lead people to a wrong conclusion. So next time on, when you are explaining a concept don’t hesitate to talk intelligently. I can assure you many people in the audience will understand what you are talking about and will appreciate it.

(Another interesting anecdote here. One of the speakers, speaking about a software solution confessed, “Earlier whenever the client had any problem, we would say “Buy more storage.” Nowadays we don’t say that.” Turns out they are asking clients to buy more hardware and more software!!)

The last observation. In every conference I notice that the registration process is generally outsourced and young girls with business formals are the ones who do your registration. It is almost become a de facto standard. It is as if the companies are afraid that if attendees do not see this happening, they will most likely say, “What, no young girls in business formals ? I don’t want to register. I will not attend this conference.” This is really not a crib but an observation. I don’t mind it actually.

Before signing off, let me be positive. The VMware conference which I had attended a few months back was very good. All the speaker spoke sensibly and did not insult the audience intelligence. Attending such conferences is a joy. Hope others take a leaf out of it.

Can you sell what you innovate?

“Cometh the hour, cometh the man.” Or woman, or in the case of the industry, ‘the machine’. Tough times are when innovative ideas are most needed but will people buy them? I post my thoughts now on the twin aspects of innovation and marketing.

This blog post of Jon Taigo led me thinking in this direction . He has a blog called ‘Drunken Data’ and for those who haven’t read his blog posts, he is one of the most vocal persons when it comes to giving his opinions on subjects which he holds close to his heart. He is not known to mince words and needless to say, his posts make interesting reading.

This post held my interest because it spoke about XIOTech’s Intelligent Storage Element (ISE). This is an innovative product. They have an innovative heal-in-place capabilities for the disk drives in their ISE. Errors are detected and potential problems are fixed before they can occur. Automatically moving data to a spare drive and rectifying a failed (or about to fail) drive saves on lengthy RAID 5 rebuild times. You can get more details about their product from their website http://www.xiotech.com

If you read Taigo’s blog, you will see statements made by prospects and competitors like, “questionable future of a company like XIOTech.” While it is true that innovative products are the need of the hour , it is also true that enterprises want to take the minimum of risk in these times.  I am not saying that competition happens only during these times but the competition is bound to be very stiff given that the total spending pie has shrunk. Added to it, the companies turn very conservative.  The innovative smaller companies may face more heat during these challenging times when it comes to selling their products. While the value the innovative product delivers may be excellent, (theoretically maybe), it would still be an untested product and more importantly from an untested company. Hence the reluctance to buy an innovative product.  When it comes to the question of a better product or a trusted company, the client may jump in favor of the latter. I don’t have statistical data to back this up but going by the reaction of many people I meet, I am guessing this would be true. These are the best of times for big companies to come out with some innovative products. The industry would lap up such products. My feeling is that companies sell themselves off or merge with other companies, not just because they want to make tons of dollars, but because brand and reputation building is a long and tough road.

I am ofcourse more concerned about the fate of people like me than about small and large companies. Naturally :). The tough times throw up multiple challenges. While everyone advocates some basic things: ‘find what problem the prospect has’, ‘find the pain areas’, ‘give a great value proposition’ etc, it is easier said than done. That is why you find lot more people saying this than doing it!! How would you investigate a pain point of a prospect if he/she is not even willing to meet you? What I have observed is that it much tougher getting a prospect to talk to you in these times. The general response would be, “We are still finalizing our strategy”, “We are in the process of finalizing our budget” etc. It is then you realize the value of your contacts and also the value of a brand name.

The only way out is to keep up the struggle and keep coming up with innovative ideas. At the same time we should be building a brand name for ourselves. We need to take help from all people we know, get a foot in the door and deliver. The last part is what is going to help in the long run. Your friends can help you get the foot inside the door. Then it is upto you to deliver. Your work is the one which will stay with you and help you or haunt you for a long time. There is no magic potion other than good old fashioned hard work combined with good old fashioned smart work. And yes, do find out that pain point of your prospect, one way or the other. That will go a long way in easing your pain as well.

A Bull and Drivers

It is not about locking bulls on to a bullock cart and driving it. It is about how you castrate a bull!! Keep reading to find out what I am talking about.

The book everyone in the Storage world is talking about currently is the book by Dave Hitz,the founder of NetApp. It has a  very eye catching title  ” How to Castrate a Bull: Unexpected Lessons on Risk, Growth, and Success in Business.”  You can read the Chapter Zero at Dave’s blog A very interesting Chapter Zero. Around the time the book was released, NetApp also topped the “Best Company to work for”  list released by Fortune Magazine. So everyone is now keen on reading  Dave’s book in order to get an insight on how NetApp created such a culture. The book has been released in US. I am not sure if it is available in India yet. I intend to buy and read this book. Will post my thoughts on the book once I read it.

I had the good fortune of meeting Dave once when he visited India in the early 2000s. I was part of  Wipro, where we hosted him for a meeting. It was pleasure interacting with him, though it was a brief interaction. He was a very down to earth person, absolutely no airs, asked questions with a real intent to learn about work done in India and was very articulate about his vision. NetApp didn’t have a development center in Bangalore then. They later started their India operations and are doing quite well.

In my last post I spoke about some storage technologies that we need to concentrate in these time. What I left unsaid was the fact that you need to have your fundamentals clear and strong. Only then will reading up and learning these new technologies help. And nothing gets more fundamental than writing device drivers.

Here comes the second part of my title. I want to recommend the book written by my friend and ex-colleague, Sreekrishnan Venkateswaran, “Essential Linux Device Drivers”. This is a welcome addition to the literature on Linux Device Drivers. The books which are commonly available here are Pajari’s, “Writing Unix Device Drivers” and the O’Reily standard, “Linux Device Drivers” by Corbet and Rubini.  Sreekrishnan’s book covers writing device drivers for a lot of devices. You will find device drivers here for I2C, PCMCIA, Blue tooth, WiFi etc. In short, it is very up to date with respect to devices that it talks about.

You will find lot of good reviews of this book on the web so I will not write a detailed review here but will talk about Sreekrishnan instead. Sreekrishnan is an IBM veteran being with IBM India for a long time. I came to know him when I joined IBM for a brief stint starting Nov 2006. It soon became apparent to me that the whole team was looking up to Sreekrishnan when it came to technical matters. He was always the first one to be called whenever there was any technical issue that needed immediate attention and whenever there was a fire to be doused. He was the also the technical face which was projected to visiting prospects,. Without exception every prospect would be impressed by Krishnan’s indepth Linux knowledge.  Krishnan and his team have ported Linux onto lot of devices including a wrist watch!! Whenever there was a client demo on our Linux porting capabilities, Sreekrishnan would show the Linux port onto a wrist watch as an example. In order to point out how feature hungry people were, Krishnan would remark, “What is the use of a Linux wrist watch if you can’t get stock market quotes on it in real time!!. So we had a implement that as well”. This would bring a smile on the face of the person watching the demo. I see that Krishan has used the same example in his preface as well, where he documents the issues which he and his team faced while porting Linux onto a wrist watch in late 1990s.

Sreekrishnan is an extremely approachable person, and, as with all good technical folks, is passionate about what he does. He is generally very happy to discuss technical matters with you and clear any doubts that you may have. The book is written in a very engaging style and makes the subject interesting. When I read the book it is almost like hearing Sreekrishnan speak.

Krishnan has been contribution regularly to Linux Magazine for quite some time now and his hands on experience in Linux clearly comes out in the book. This book was published by Prentice Hall last year in US under its Open Source Development series. This book is now available in India as a low price edition. I would definitely recommend anyone working in the Linux area and is writing  (or intends to write) a device driver to get this book. It is definitely worth keeping this in your reference library.

Nothing enriches your life like a good book. Here is a lovely quote I read about books. There was a person (I forgot his name) who is supposed to have bought lots of books and built a massive personal library. When someone asked him why he spent so much money on books, he replied, “If I had not bought these books, I would have had lot of money. But I would not have been richer”. I vehemently agree with him. (In case someone knows who made this quote, let me know. I will attribute it accordingly.)

I will leave you to mull on that quote.

Technologies for these troubled times

“What technologies should we  come up to speed in these troubled times”, is a question that students and engineers freshly into a job ask me. Times are tough, especially for guys coming out of college and those having a job in hand but  not working on client  project yet. Here is my take on some recent technologies areas which everyone wanting to work in the server and storage domain should know about.

Virtualization is something which will be taken for granted a few years from now. Every enterprise would have implemented some form of virtualization or other. Virtualization, in case you didn’t know, is the act of showing something that doesn’t exist. Some people feel Ramalinga Raju of  Satyam was a master in this. The only difference is that the virtualization we are talking about is beneficial to the user whereas the sort of virtualization done by Enron, Madoff and others of their ilk is beneficial only to the ‘virtualizer’.

In case anyone wants to get on the Server Virtualization bandwagon, it is quite easy.Free VMware player is available for download (http://www.vmware.com/products/player) .  I downloaded VMware player and use Ubuntu Linux in it. That enables me to write some programs in the Linux environment while still using my Windows to check my mails and prepare Powerpoint slides.  (Who can ever escape Powerpoint?) In VMware marketplace there are many virtual appliances available for free. (http://www.vmware.com/appliances/marketplace.html) If you have the free disk space and time on your hands, you can download them and try it out. This will give you an idea about server virtualization. If you love Open Source, you can try out Xen Hypervisor. I have not tried it out yet but will do so soon. (www.xen.org)

In the Storage area, assuming you have a basic understanding of Storage technologies, Thin Provisioning is an technique which you can read up on. This is a technique wherein you show the user a lot of disk space as his quota but you don’t need to have all that space physically available right now. (Think  Satyam again) You keep buying storage as the users keep filling up capacity. This is something which all free email service providers like Google, Yahoo, Rediff etc would be doing. It makes no sense to buy all the advertised storage. (I mean, how can you buy ‘unlimited’?) The companies would be buying more and more storage as your mail account keeps growing with all those mp3 and jpeg files. This technique is called Thin Provisioning and the key aspects to achieve success in this are capacity planning and capacity forecasting. So read and understand how Thin Provisioning works and how it is implemented. I am sure if you are in the Storage array space you will have to do this sooner than later.

The next Storage technology you may want to look at is Data Deduplication. This is the simple idea of removing all duplicate data by storing a single image and pointing all duplicates to this image. There are multiple ways to do this and there is debate on where it needs to be done. Will post on this debate at a later date. Virtual Tape Library (VTL) solutions come with Data Deduplication. So here is something you should be looking at. Even in these tough times Data Domain, a company focused on Data Deduplication has posted impressive results.

Another technology you can read up on is Wide Striping and Micro RAID. Rather than me explain I will point you to this lovely video by Mark Farley of 3Par, where he clearly explains these concepts. (BTW, Marc makes most of the videos you seen on his blog while driving!! He uses a steering wheel cam. This is one of the few videos he has made when he wasn’t driving. Shows how quality of improves dramatically when you feet are firmly planted on the ground 🙂

Things like Solid State Drives and Fibre Channel Over Ethernet (FCOE) are happening on the infrastructure front but as engineers and system integrators concentrate on virtualization and the other technologies that I mentioned. Our role in Solid State Drives and FCOE will be limited in the immediate future.

Before I end, I need to say something to all college grads which should actually be unsaid. But I will say it anyway. More than reading up all this stuff, ensure your basics are strong. Your programming skills, your understanding of the operating system and your analytical ability are things that anyone planning to hire you will look for initially. So strengthen your basic by working hard on them regularly. Now that I have given this piece of advice, I can sign off  in peace.

(I got an indication on how bad times are for all industries when the Toyota service centre folks called me on their own and give you a service  appointment. “Whatever day and time  you want, sir”. Made me realize the enormity of the problem)

Update:  I had not checked my Google Reader for a couple of days. After posting this I checked my reader to realize that Marc of 3Par has now put up a video explaining chunklets. Check it out.

Radhakrishna Bairy – A small tribute

New Year dawned on a sad and tragic note with the passing away of a friend, former colleague and a great human being, Radhakrishna Bairy. It was shocking to all of us that Bairy, as he was fondly known, went for a mundane hernia operation and never came back alive. No one fully knows nor can comprehend why this tragedy happened.

Bairy and I worked together as part of the Sequent (later IBM) offshore account. He managed the ptx/Dynix team. I had met him a week before he went for the operation and we were discussing various issues, personal and professional, for quite sometime. Even now I find it difficult to believe that he is no more.

I had great respect for Bairy both on the professional and personal front.  As a people manager, he had the knack of assessing the talent of each of his team mates perfectly. I can say without contradiction, that all those he rated very highly have done very well professionally. They are  in good positions in various companies. He always gave me an exact idea of the capability of his team members and he never let any personal issues with the concerned person affect his assessment of the person. In this aspect, he was a professional to the core.

His sense of work ethic  was impeecable. He was a person who knew his own limitation and would readily admit it. He was one you wanted on your team since he was always positive about any new program that you wanted to try out. He would suggest ways in which  it can be done better and he would work hard to make it a success. Any task assigned to him got his full attention and he was always thinking of ways to effectively complete the task.

More than the professional aspect, it is in the personal sphere that I wish I could emulate one of his key characteristics . His total lack of bitterness, whatever be the situation. He was a very straightforward person and told you what he thought about any subject. He was always open about his disagreement, if any. He would express his unhappiness with people but there would never be any bitterness associated with it. This is a very rare quality and you can see it only in very few people.

All our hearts go for Lalitha and the children, who have to face this mammoth tragedy. I sincerely pray to God that they can overcome this tragedy in due course of time.

Bairy will be missed by all of us. May his soul rest in peace.

It's virtual everywhere

In my last post I had spoken about the VMware virtualization seminar. Coincidentaly I had to do some work on virtualization for an important client of mine. As I was going through lot of literature related to virtualization, what stuck me is the inroads that virtualization is making into multiple spheres. Ofcourse I have always been a server and storage guy and looked at virtualization as something more related these areas. It was quite instructive to learn that virtualization is being used in various other areas as well.

Take the case of networking. I was aware of Virtual LAN(VLAN) and Link Aggregation. I now found out that Cisco has introduced switches that virtualize across switches. That is, they can take two switches, connect them together and show them as a single switch on the network !! User see more ports, it provides reliability and these switches share their routing tables. Of course, Cisco had also implemented the concept of Virtual SAN (VSAN) in their switches earlier.

More interesting is the fact that virtualization is coming on to the mobile. It is going to take some time getting there but efforts are on. ARM is supposed to be providing support for virtualization in its processors and companies like VMware are very serious about providing a Hypervisor for a mobile phone. Individual applications can run in an isolated VM so that when the application crashes, it affects only that particular VM and doesn’t hang your phone. There are challeneges ahead, especially in terms of providing real time response for some applications, power management etc.  I am sure these challenges will be addressed and in a few years time we will see phones with hypervisors.

The coming years are definitely going to see more virtualization in multiple spheres. In cases of servers, I think virtualization is a logical thing to do. I don’t mean from the point of view of consolidating servers, saving costs etc., but as a logical thing to do.  I mean, do you think anyone would buy multiple TVs just because each TV can only show you certain channels? We don’t seem to blink when it comes to running applications. We buy multiple systems just because an application will run only on a certain platform. Since the basic function of a computer is to run an application, worrying about the application should be the priority and not worrying about the underlying OS. Virtualization helps achieve it. It will be too much to ask for a single OS for all systems, isn’t it?

Read an interesting article on how Storage is spreading to different areas. This is an article about Storage usage by Avid, the video editing folks. You can read the article here. A good friend of mine works at Avid. It is instructive to read this article, especially the last couple of paragraphs. It shows how much we take the customer requirement for granted and maybe never ask the customer what they need !!

It’s the year end now and as usual there are lot of articles on 10 Best Whatever of the year. To us Storage folks, we need to follow what is happening in the Industry. This article gives the Top 10 Storage acquisitions that happened in 2008. No doubt this will affect our future as well going ahead.

Wish you all Merry Christmas and a Happy New Year 2008. I will be out after Christmas on a short vacation. Catch up with you next year.

VMware seminar and some thoughts on virtualization

On Wednesday I attended a VMware Virtualization seminar at Hotel Chancey Pavillion, Bangalore. It was a well attended, half-a -day, seminar. It must have been a combination of VMware’s popularity and the current market scenario, that brought in a large crowd and the huge hall was full. The other surprising thing was that most people arrived on time and the seminar started at 9 am as announced. A rarity, which the first speaker made a point of mentioning.

The seminar was done professionally. Things started on time, the speakers spoke well, the presentations were brief and to the point and importantly all speakers finished their presentations within the alloted time. In short, they ensured that you were not bored. A major achievement for a seminar !!

In most of the seminars sponsored by the marketing teams, the theme always remains the same, “We have the exact solution to every problem that you have.” You need to observe what they are not saying or don’t want to say in order to get a more realistic picture. This seminar spoke about the future features and products from VMware. From what I could gather from the presentation, the two key areas they are keen on addressing are:

  • Running business critical application in virtualized environment
  • Management tools to support server virtualization

They had a presentation about running critical apps like Exchanger server, SAP, Oracle in virtual machines and showed how this could enable better performance. Given that VMware would want to be in the heart of the enterprise, addressing this issue will be of paramount importance.

Management of virtual machines is another headache which requires lot of good tools. Currently VMware has certain tools and they are quickly adding to this list. The future tools, as always, give an idea of what is missing now 🙂

The speaker on Desktop Virtualization was very gung-ho about it. Reminded me of the X-server days when we thought that the diskless workstations running X-servers would catch up soon. Somehow that didn’t happen to the degree we anticipated. It will be interesting to see how much Desktop Virtualization catches on. To me the adoption of Desktop virtualization would be slower than Server virtualization but a good value proposition exists.

VMware folks say that they have around 600 customers in India. I am sure this number will increase soon. In a country like ours, virtualization makes great sense. Especially given the power shortage and the current quality of power. This was emphasized in the seminar and I fully agree. The lesser we draw power the better it is for everyone. Maybe we should make virtualization mandatory for companies which draw more than a certain amount of power for their IT needs. I am sure the server vendors will not be very happy whereas the virtualization guys will be more than happy. Given the value proposition of virtualization, it doesn’t need a law for people to see the benefits. Especially in the current economic condition.

A nice article by Eric Seibert, which talks about what you must do after you have virtualized. Read it here.

Met a manager from a company which implemented virtualization. He was saying how difficult it was to get expert help. The company from whom they bought the infrastructure would send some ‘experts’ initially for deployment. After a few days, they would be pulled out and new faces will appear. Upon continuous pressure the old guys will return but only for a few days. When they tried getting some consultancy from the software guys, their quote for consultancy services was more than the cost of the infrastructure !! Apparently they had their experts in Singapore, who had to fly down and stay in some fancy seven start hotel for which this company had to foot the bill !!! This is something I have personally seen, being on both sides of the table. Consultancy and training are high margin stuff and companies want to do this on their own but unfortunately do not have enough staff to meet all the demands.

Cloud(ed) Management

It is usual that whenever newer technologies, techniques or paradigms emerge, there is a certain amount of confusion. On one hand these are generally held as a panacea for all ills that plague us. On the other hand, it is felt that this wouldn’t work for some reason or the other!! The truth generally lies somewhere in between. I remember the hype we had in my former company when we started work on a Bluetooth stack. At that time it was felt that every connection in the future would be Bluetooth. The hype died down and the team was disbanded before the world discovered the actual uses of Bluetooth !! Well, talk about getting the timing right !!

I think Cloud computing and Cloud storage are going through a similar phase now. You need to look into all sides of the argument to bet on how this will pan out in the near future. Thanks to all the various blogs by people from different companies and other articles, users are getting to hear multiple arguments.

Dave Graham, of EMC, has started a series of articles about Cloud Optimized Storage. I plan to follow them in order to get a good understanding of what this technology is all about. You too can read the article here and follow Dave’s future articles.

There was an interesting blog by Martin Glassborow, a Storage user, whose blog is titled ‘Storagebod’. He had a post which pointed out the lack of good management tools for managing storage. You can read his post here. A hard hitting post which elicited good responses with people sharing their storage management woes.

This leads to me thinking on how on earth are the Cloud companies going to manage their storage !! The Cloud company is going to manage multiple data centers. The cost effectiveness for the Cloud company comes from consolidating lot of storage and allocating storage in an optimized fashion. This is easier said that done. If the current tools do not allow for effective management of a data center of one single company, imagine what will happen when data centers of many companies are merged together in the cloud!! Things become even more complex when server virtualization is added to this mix. It will be interesting to see if any specific tools to manage the cloud will emerge or if the administrators of the cloud will have to do with the tools available currently. I am sure no one envies the administrators of the cloud.

An argument against the Cloud. Here is an article by Andi Mann, which questions the cloud from a security and compliance perspective.

Interesting times indeed. Interesting and tough. We will talk about the tough part in the next post.

Castles in the Cloud

I will try and post the recent happenings in the Storage world in my blog every week. Atleast I will try 🙂 In this post let’s talk about the current hot topic, the Storage Cloud.

The idea of Storage cloud is to enable you to build your Storage Castles in the Cloud. The idea is that data will no longer be stored in your own data center but will be stored online elsewhere. You can say that this elsewhere is what constitutes the Cloud. This Cloud would be run by a company which will provide this service of storing the data and would enable you to access your data as fast as you would in your own data center. In essence, the data center operations of your company will now be outsourced to this ‘Cloud company’. Your enterprise will need less number of admins, less management and supposedly less headache. It is the duty of the ‘cloud company’ to deliver the storage to you as per the agreed upon SLAs. Of course, you need to pay for this service. And probably pay a lot.

Sounds simple enough, doesn’t it. As they say, every complex problem has a simple solution which is wrong !! Not that the concept of Cloud Storage or Cloud Computing is wrong. It is just that there are lot of questions being asked. Like, how much will this model affect the networking bandwidth, will such a model be ok from the compliance point of view and so on. Google and Amazon are two companies who are generally quoted when referring to ‘The Cloud’ and both of them had outages recently. So the debate is on.

Why would the Cloud succeed when a similar concept like Software As A Service (SaaS) did not take off? (Many of you would remember about this buzzword during the early part of this century.) That’s the million dollar question. Or shall I say the multi-million dollar question. My guess is that for a couple of reasons Cloud may turn out to be different from SaaS. One, the data growth is tremendous and the complexity of managing data is also increasing. So many companies may want to outsource this activity to the experts rather than keeping a large IT team and buy more and more storage. Second, lot of the big guys have entered the fray. Google and Amazon have shown that cloud can work. Microsoft recently announced an offering called ‘Azure’ targeted at the cloud. EMC announced Atmos, short for Atmosphere, which it calls as Cloud Optimized Storage. (COS) (We will have to wait and see if the industry will pick up this term.) The combined marketing strength of these companies has already created a lot of buzz about Cloud, thereby increasing the prospects of its adoption by the industry.

You can read about EMC’s Atmos in this blog by Storagezilla, an EMC Blogger. As you can see, Atmos is designed for the cloud, information being stored as objects and policies which act on these objects. Policies are also used to drive geographical data placement. Initially, the reaction to this announcement from EMC’s competitors was that it was nothing new and not innovative enough. Slowly I see a perceptible change and people saying, “Oh yea, we too have it !!” This probably means that EMC is onto something big and saw the market before others did. As usual, time will tell us how successful this will be but it has surely created a buzz in the Storage circles. Companies like EMC will not invest in coming out with such a product if they didn’t believe Cloud would work.

What does all this mean to guys like us who thrive on services business? We need to learn atleast three things if we need to succeed in this area. One, about the Cloud itself. Second, about Server Virtualization, since the Cloud will pretty much be virtualized. Three, Storage virtualization. All three will be important in our business.

That’s it for my first technical post. Will continue with more updates about new releases and more action from the Storage land soon.