I was invited to participate in the New Jersey Digital Government Summit in Trenton, New Jersey, on May 8.  One of the most interesting sessions was titled: Situation Room: Hurricane hits New Jersey! This was chaired by Jack Mortimer of Government Technology magazine, and participants were Gloria Broeker, Statewide Disaster Recovery Planning Officer for IT, State of New Jersey and Craig Fugate, Director of Emergency Management, State of Florida. This session was modeled on the concept of the White House Situation Room, and took the audience through a simulation of a category 4-5 hurricane approaching and finally hitting New Jersey. Jack did an excellent job, taking the audience through the different stages approaching the landfall, posing questions to the audience which consisted of New Jerseys IT organizations, and engaging Gloria and Craig as the panel of experts.

Craig, being from Florida, had a great deal of experience with hurricanes and gave first hand accounts of what needs to be done going up to, through the event, and cleaning up afterwards. Gloria, who has responsibility for Disaster Recover for the IT systems in New Jersey, talked about her planning efforts for disasters in general.

A category 4-5 hurricane has not hit New Jersey but that is not to say it won’t happen. Jack quoted a forecast by Klotzbach and Gray which predicts that the chances of a major hurricane hitting New York and Long Island have increased to 24% in 2008 from an average of 16%. The chances of a hurricane category 3-4-5 hitting the East Coast of the US has increased to 45% in 2008 from an average of 31% due to warming water temperatures and other climatic factors. These probabilities were not available for New Jersey, but they are right next door to New York.

There were many new things that came out of this session.

While data centers in this area have diesel generators to maintain power during a disaster, many are not prepared for the loss of water for the chillers. When a hurricane comes through and rips up the trees it also rips up the water mains that are tangled in the roots of the trees. In California, I would assume that we would have a similar exposure in earthquakes. We need to store water as well as diesel fuel for the generators.

When a wide scale disaster hits, the first priority of the State is not to recover the servers and storage and get the IT up and running. The first priority is to establish communications which usually means satellites since cell towers and land lines will be destroyed. When Katrina hit there was no communication to the area until television crews were able to get in almost 48 hours after. Second priority is to establish security so that desperate people aren’t stopping rescue and recovery efforts. The next priority is to rescue and recover the people. After that you recover your IT. 

In the planning process before the disaster hits, determine who will pay. No one has adequately budgeted for this, but you do not want to be caught in a paralysis trying to make this decision when a hurricane is barreling down at you at 50 miles per hour.

When such a disaster strikes it affects everyone at the same time in a wide area. If you out source to a DR Service realize that they will be swamped. Who will they service first? Even if you are outside of the immediate area, your power grid may be out and diesel for your generators may not be available for 48 hours or more. Make sure your vendors and suppliers are alerted to your impending need for support.  

Build recovery teams to respond to problems you never imagined. Craig talked about the need for being “super Gumby”, be extremely flexible.  “Have a plan but realize that disasters never go according to plan. If everything goes according to plan it is an exercise” 

We all left that session with a heighten awareness and appreciation for disaster preparedness.  One of the last questions asked was how many in the audience had a disaster preparedness plan for their family. Unfortunately, this response was much lower that the response for disaster preparedness for work.

Craig has a website disatersrus.org which offers plans, alerts, and links to international sites with similar news and information about disaster preparedness.          

How do you buy a storage system? Not too many years ago you probably bought a fully configured, storage system, and depreciated it over 3 years. You only needed less than a third of the capacity when you bought it, but capacity was cheap and you figured that your workload would increase and you would fill it up over the next three years. Even though you knew that the cost of storage capacity erodes quickly over time, you bought more than you needed at current prices because it would be too disruptive to add the capacity in increments.  Besides the vendor gave you a great deal, and you would quickly fill it up anyway.

At the end of three years you found that the capacity of a disk is now four times what it was three years ago for the same price. However, to use these disks you need to upgrade to the next generation storage system with higher speed Fibre Channel ports, larger cache, new replication functions, and more than double the number of disks. Buying a fully configured system for your next round is out of the question. So this time you, decide to take a more measured approach. You buy only the control unit and enough capacity to cover the next year’s growth, with the thought that you can add capacity as you need it at a lower cost in the future. A year later when you go to add the additional capacity, you find you need to upgrade your control unit with more ports and cache as well as more storage frames to hold the additional capacity. You also find that the cost per GB is cheaper but it comes on fewer disks. Since you need a certain number of disks for performance, you buy more capacity than you need. An outage is required to add the new capacity and reconfigure the cache and port assignments. A year later you repeat this again for the next year’s storage growth.

Now at the end of three years, you find that the original control unit and capacity that you bought three years ago are no longer under warranty and you have to start paying maintenance fees which are quite expensive. In fact it is almost cheaper to buy the next generation control unit and the larger capacity disks, than pay maintenance on the old control unit with its three year old disks. But wait, if you swap out the control unit, you also have to swap out the capacity that you just bought over the last two years, since the control unit and disk capacity are bound together. So you either pay the maintenance charges or you give up the depreciation on the upgrades you purchased over the last two years.

You would not have to do this if you could separate the control unit from the storage media and capitalize them separately. With the USP VM, you can buy the latest high function storage control unit without buying any storage media. You can connect it to your existing storage systems and upgrade the functionality of those storage systems with the latest USP storage services like volume virtualization, Dynamic Provisioning (thin provisioning), universal replication, and non-disruptive migration. If you need more capacity you can buy another storage system with the latest large capacity disks and connect it to the USP VM. With Dynamic Provisioning you only need to buy the capacity you need since the new disks can be added to a dynamic storage pool and the I/O will be spread across all the disks in the pool to level out the spindle load. When it comes time to replace the USP VM, you can continue to depreciate the storage system that you bought later by attaching it to the next generation USP V or VM.

The separation of the storage controller from the storage media is the next step in the disaggregation of storage. In the beginning, storage and servers were integrated. When you needed to change a server you also had to change the storage. That problem was solved by separating the server and storage. By separating the storage controller from the media, you now have the choice of mixing and matching controller functions and storage media to provide the best solution for your storage requirements. And with storage virtualization and dynamic provisioning, you only need to buy what you need when you need it with out disruption to your applications.

Unlike other thin provisioning solutions, HDP, or Hitachi Dynamic Provisioning, supports other vendor’s storage that are attached, or virtualized, behind the USP V or VM. Since it is provided as a service on the USP V/VM, it can aggregate thin provisioning with other common storage services provided on this platform. This Services Oriented Approach to HDP not only provides thin provisioning to other vendor’s legacy or lower level storage system through storage virtualization, we can also combine this with other USP V/VM services like snapshot copies, non disruptive moves or migration, distance replication, and Copy on Write and enhance the capabilities of these services. 

With HDP we provision 42 MB pages of virtualized storage as a user actually writes to his allocated storage space. When he exceeds his 42 MB page, we give him more pages until he reaches his maximum allocation. If his application enables him to increase his initial allocation request, we can continue to add additional pages. 

Since the USP V/VM knows how many pages HDP has provisioned to a volume, when a copy is requested, the USP V/VM only copies the pages that are provisioned and not the entire volume. Without HDP a copy function would have to copy the entire volume, including the allocated unused space, since it would have no idea how much of that volume was actually used. This could provide a tremendous saving in allocated unused storage capacity when one considers how many copies are made for offline back up cycles, replication, development test, data mining, data distribution, etc. Without HDP, all those copies would replicate that allocated unused space, over and over again. 

HDP can also be used with dynamic moves for device migration at end of lease, tiered storage for performance tuning or cost reduction, and consolidation following a merger or acquisition. Since only the used pages and not the whole volumes are moved, the moves can be made more efficiently. 

This same efficiency applies to synchronous and asynchronous replication and to Copy on Write. Replicating and recoverying active pages rather than allocated volumes will be more efficient and less wasteful of capacity.

This combination of efficencies can only be realized when you take a Services Oriented Approach to thin provisioning, rather than a stand alone approach.  

Since I posted my blog this morning on Earth Day, EPA and PUE. I received some pictures of the Green data center that Hitach is building in Yokohama. You can check them out on flickr. This should be on line by April 2009 and is targeted for a Power Usage Efficiency of 1.6 which will be one of the lowest for a Tier IV data center.

SUN has already built a Green data center in Santa Clara, which I was able to visit with our SUN partners yesterday. They converted a slab floor office space into a Tier I data center in about a year’s time. This data center has a PUE of 1.28, which I would think is the lowest even for a Tier I! They have done some innovative things which they hilight in  a video on their website. Their different approaches to cooling racks was very interesting. They give breifings and tours, which you can schedule through your SUN representatives when you are in Santa Clara. For information about the SUN PUE rating, you can link to Dean Nelson’s blog. Dean is the Director of Global Lab and Data Center Design Services. 

Next year, about this time,  we will be able to show you Hitachi’s Green Data Center when you are in the Tokyo area. 

Earth Day 2008 will be celebrated on April 22 with observations scheduled around the world. This year’s theme is “A Call for Climate” a global warming action theme. While there may still be skeptics about global warning, there is no disputing the rising cost of energy, as I fill my Prius with $3.89 per gallon gas in California.  

In IT, the increasing demand for power and the cost of that power has quickly emerged as primary concerns, driving companies to take a closer look at the efficiencies of their data centers and of the IT infrastructures that they support. The United States Environmental Protection Agency (EPA) presented a report last August to the United States Congress which focused on IT data center energy usage and associated issues. This report noted that while data centers only accounted for 1.5% of energy consumption in the United States in 2006, they were a vital national asset and any disruption in availability of services would have severe consequences for business as well as national interests.  The report also estimated that this consumption would double by 2011. 

This report surveyed a number of data centers to track growth from 2000 to 2006 and found that 50% of the energy went into heating, ventilation, air conditioning (HVAC), power conditioning and distribution. The rest of the energy went into servers, storage, and network equipment. This was not surprising since most data centers were built 10 or more years ago. The last big data center construction boom occurred during the dot com boom in the late 1990’s. While energy costs were much lower then, we did experience energy shortages and brown outs in California where many service providers were located.  

As a result many companies will be building new energy efficient data centers over the next few years. In order to measure the efficiency of these new data centers, The Green Grid consortium has defined a data center power efficiency metric called PUE for Power Usage Efficiency. 

PUE is calculated by dividing the total power usage of a data center by the total power usage of the IT equipment. According to The Green Grid that ratio may be as high as 3.0, meaning for every watt that goes into IT equipment; 2 watts go into the data center infrastructure. The EPA study would suggest that this was closer to 2.0. PUE will also vary according to Data Center Tier, where Tier IV is the highest tier which supplies active/active redundant power supplies and supports less than 0.4 hours of annual IT down time due to site failure.  

As part of their CoolCenter50 initiative to reduce data center power 50%  in 5 years, Hitachi Ltd. issued a statement that they have set an aggressive PUE target of 1.6 for their new class IV data center in Yokohama, Japan. This data center is scheduled to be in production by April 2009 and will be architected to leverage the collective product innovations of the Hitachi Information and Technology Systems Group (ITSG) which includes storage systems, servers, and networking equipment and other key areas of Hitachi technology in thermal hydraulic cooling systems, uninterruptible power supplies, and power supply converters.  

While PUE is a relatively new metric, it is important to set targets and agree on metrics to measure progress in addressing energy as well as environmental issues. As we build out this data center, Hitachi expects to learn many lessons which we can apply to these issues.       

This was the title of a panel that was held last week at SNW in Orlando. It was moderated by Ron Milton of Computer World and on the panel was Mark Showers the CIO of Monsanto, Mark O’Gara Vice President of Infrastructure Management at Highmark, Andrew Fanara, team leader of the EPA Energy Star Product Specification Development Group, Chris Wood the CTO for SUN Global Storage Practice and myself.

At the beginning of the session, the question of why I should care about having a “Green” data center was posed to the audience, and the top two responses were;

1.      Reduce costs, improve company profits

2.      Protect the environment

This clearly showed that the majority of attendees clearly believed that this was a journey to a new reality. Andrew Fanara pointed out that this could be motivated by the reality that we are running out of power. The recent trend of relocating data centers to the Columbia River area for lower cost hydroelectric power is causing a shortage of power in that region. Building new energy production facilities will take time, which means we need to conserve what we have today. 

Both Highmark and Monsanto have built new data centers that have LEED certification, Leadership in Energy and Environmental Design, which is a green building rating system developed by the U.S. Green Building Council. Mark Showers and Mark O’Gara emphasized the value of Green to their IT mission. Both talked about the need to establish a base line measurement in order to know what to address first in terms of power consumption.

Later I had the opportunity to have lunch with Mark O’Gara. Mark is a West Point graduate so he takes a very disciplined approach to addressing the greening of IT. He emphasized the need for measurements and setting targets. When he started out he did an analysis of power consumption based on vendor specifications and came up with a number of 513 KW for his data center infrastructure. When his infrastructure team actually measured the power consumption, the peak draw was 462 KW, about 10% less than the plate KW. It is good news that vendors over spec rather than under spec their products to ensure that facilities’ planning for power is never exceeded. However to get an accurate reading of power consumption it should be measured under peak conditions on the floor.  

The physical measurements showed that the biggest consumers of power were in order: Business Intelligence Servers, SAN Storage, Robotic tape Library, and Virtual tape servers. He also found a lot of systems that were not being used. His first step was to turn off those servers and move to virtual servers. Mark’s initial target was a 10% reduction of power and he was able to achieve a reduction to 418 KW.

At first it may sound counter intuitive that SAN storage is such a high consumer of power since it is open system storage which is usually rack mounted as opposed to large monolithic mainframe storage. However, rack mounted, modular storage is very similar to rack mounted servers when it comes to power and cooling. Since it is modular each drawer of 15 disks has to have its own power step down and two or three fans for cooling. Cool air has to be sucked in from the front and exhausted out the back, creating a ventilation problem on the backside. On a capacity basis monolithic storage may be more power and cooling efficient since they can be configured in frames with a wall of disks in front and a wall of disks in back with a plenum in between which acts as a chimney to draw air in from the bottom and sides and exhaust hot air out the top. So instead of 2 or 3 fans per 15 disk drawer, you may have only six fans for 256 disks in a monolithic storage frame.

Another surprise may be that tape libraries are such large consumers of power. Since tape is not spinning most of the time they should consume much less power than spinning disk - right? Apparently not if they are sitting in a robotic tape library with a lot of mechanical moving parts and tape drives that have to accelerate and decelerate at tremendous speeds. A Virtual Tape Library with de-duplication factor of 25:1 and large capacity disks may draw significantly less power than a robotic tape library for a given amount of capacity.

The greening of IT is a new reality. The lessons learned by thought leaders like Mark Showers and Mark O’Gara is to measure what you have first so that you can identify and prioritize an approach to reducing power and cooling requirements.  

There are multiple dimensions to Hitachi’s implementation of dynamic provisioning. Hitachi Dynamic Provisioning, HDP, is software that runs on the USP V and USP VM storage Controllers and provides the ability to dynamically provision storage as it is required from a pool of storage that is wide striped across any number of logical devices or RAID sets. The available capacity in this storage pool is closely monitored and soft and hard alerts are given when a dynamically provisioned volume approaches its maximum allocation. Alerts are also given when the available capacity in the pool approaches predefined high water marks. When these alerts are triggered, additional capacity can be added to the pool or a volume can be on non-disruptively migrated to a larger pool.

Since a common storage management practice in open systems is to over allocate more storage than is expected to be used, HDP can be used to satisfy the allocation request by thinly provisioning the storage capacity in page stripes of 42 MB as the space is written up to the original allocation request. The rest of the allocated unused space can be used by other applications that share the dynamically provisioned pool. Some file systems and data bases that do a “hard” format when the space is allocated will not benefit from thin provisioning. However they can benefit from other aspects of dynamic provisioning, like the ability to add capacity dynamically to the pool instead of having to install and format another RAID set. Hitachi Data Systems has a number of best practices white papers on HDP, including best practices for Oracle and Windows at http://www.hds.com/products/storage-software/hitachi-dynamic-provisioning.html

Since HDP runs on a massively scalable enterprise storage system, it can be combined with other storage services like point in time copies (ShadowImage), dynamic moves (Tiered Storage Manager), Copy on Write, synchronous replication (TrueCopy), and Asynchronous replication (Hitachi Universal Replicator). The waste of allocated unused capacity is compounded many times over due to the many copies of that capacity that must be generated for: backup cycles, business continuance, development/test, data distribution, Extract/Transform/Load for data mining. HDP can eliminate that additional waste by copying or moving only the dynamically provisioned pages of the volume and not the allocated unused portion of the volume. HDP can copy or move dynamically provisioned volumes (dpvol) in one pool to another dpvol in the same or different HDP pool. It can also copy or move to a regular fixed volume in which case the volume will be fixed from then on. If a regular volume is copied or moved to a dpvol, the entire volume will be allocated in the HDP pool since the USP V/VM has no way of knowing how much of the allocation is actually used.

Even if a volume may not be able to use thin provisioning, volumes can still benefit from the wide stripe pages in the HDP pool. The basic element of dynamic provisioning is a 42 MB page that is striped across all the spindles in the pool. This wide striping improves throughput and response time by parallelizing the I/O across all the spindles in the pool. The Enterprise Strategy Group did an evaluation of HDP’s wide striping performance using a random read/write work load, comparing I/O against one array group of four spindles (2+2) versus 8 array groups of 32 spindles.  The results showed an improvement of 716% more transactions at 15 ms or less and an improvement in response time of 118% at the first data point in the curve.

 The last point is that HDP can leverage the storage virtualization capability of the USP V/M and can be extended as a service to heterogeneous storage systems that are attached to the USP V/VM. David Marshal who blogs on virtualization reports that Gartner recommends including at least one storage system that offers thin provisioning on any short list. With the USP V/VM you can install one storage system and enhance your existing FC storage systems with the ability to do dynamic provisioning, thin provisioning, and wide striping.

Today, March 29 is Earth Hour Day, where we will all be encouraged to turn off our lights and non essential appliances from 8pm to 9pm to promote conservation. The World Wildlife foundation launched the idea last year in Australia and it has now gone international.  

If you went on Google today you would have seen a black screen as their way of promoting Earth Hour. Already we are seeing the results of this coming in over the web this morning to California from New Zealand and Australia. Christchurch, New Zealand, kicked it off with a countdown in Cathedral Square. The Christchurch press reports that the combined efforts of residents, businesses and communities during their Earth hour resulted in a cut in consumption of 13.1%.during that hour. While they acknowledge that this is a drop in the ocean, it raised awareness of energy consumption and how simple it is to make savings. In the US, special Earth Hour events will be held in flag ship cities like Atlanta, Chicago, Phoenix, and San Francisco. The decorative lights on the Golden Gate Bridge will be shut off during that hour.  

Not all the comments on Earth Hour are positive Joe Mendelson in his article in the STAR, “Earth Hour: A new age, tooth fairy tale illusion”, had this comment:

I don’t think Earth Hour will make a difference whatsoever in the scheme of things. It’s delusional. It will not make people more aware of the problem. It’s a feeble recognition of our decline, and it won’t amount to a row of beans unless we plant beans every day.

I will be joining in this effort at home, but I agree with Mr. Mendelson. It is not going to amount to anything unless we sustain it.

When it comes to sustainable technology I often quote Ted Samson’s blog on sustainable IT.

“The idea is to not just haphazardly buy the newest, greenest products out there and stick them in your server rooms and on your desktops in a frantic effort to become green. Rather, it reflects planning and investing in a technology infrastructure that will serve your company’s needs today and tomorrow, while helping your company save money on wasted resources such as energy and paper; make the best possible use of existing datacenter space so as to postpone having to build a new one; and reduce its overall environmental impact, which is both socially responsible and potentially advantageous should the government start cracking down on carbon emissions and the like”.

As I go around visiting different companies I see major efforts to address sustainability.

Many companies like BP have gotten LEED certification for their data center buildings. Companies like Hitachi and Wal-Mart have adopted ISO 140001 compliance for sustainability through Energy Management Systems that cover product design, raw materials, manufacturing, packaging and distribution, use/reuse/maintenance and end of life. Companies are beginning to ask about Green Initiatives in their RFIs.

In April I have been asked to participate in several conferences on the topic of Green. Next Tuesday, I will be speaking at Data Center World Spring 2008 Conference on Eco-friendly data centers during the keynote at 9:00am on April 1. I will also be attending SNW, April 7 to 10 where there will be a number of sessions addressing Green. SNIA will sponsor a tutorial track devoted to Green Storage. Sol Squire from Data Islandia who I have blogged about before, will be presenting in a session at 1110 am on Monday April 7 on best practices and technical considerations in building a Green Data Center.

On April 8 at 1130am, I will be privileged to be on a panel titled “ The Greening of IT: Oxymoron or Journey to a New Reality” with Moderator: Ron Milton, Executive Vice President, Computerworld  The panelists include:
Andrew Fanara, Team Leader, ENERGY STAR® Product Specifications Development Group, USEPA
Mark O’Gara, Vice President, Infrastructure Management, Highmark, Inc.
Mark Showers, Chief Information Officer, Monsanto Company
Leighton C. “Chris” Wood, Jr., Director and Chief  Technology Officer, Global Storage Practice, Sun Microsystems

While I have worked with Chris Wood before in my days at IBM and through our partnership with SUN, I only know the other gentlemen by reputation. Mark O’Gara and Highmark were recognized as the TOP Green IT organization by Computerworld in February 2008, and Mark Showers was recognized as one of the 100 premier Leaders in IT by Computer World Magazine in 2007 not only for his work at Monsanto but also for his work in the community and on advisory boards for different universities. Mr Fanara will also be very interesting to meet and hear since he is currently leading the effort to develop an Energy Star specification for servers.

I hope to see you at these events, and turn off those lights and power down your Laptops when you leave the room!

Information is like beauty, in that it is in the eye of the beholder.  The same piece of data may have different value or meaning to different people who view it or when viewed in relation to other data. While HDS is in the business of providing infrastructure to store, access, and preserve data there are a whole lot of other challenges around finding and interpreting the data that we store.  

Last week I had the pleasure to introduce an exciting panel of speakers at the InfoLab/Hitachi workshop that was sponsored by the Stanford Computer Forum and Hitachi Ltd. The speakers included:

Paul Strong, Distinguished Research Scientist, eBay Research Labs
“Why Enterprise Management Is All About Relationships And Not Things”

Brewster Kahle, Director and Co-Founder, Internet Archive
“Universal Access to All Human Knowledge”

Stewart Butterfield, Co-Founder of Flickr
“Kissing Clouds”

Prabhakar Raghavan, Head of Yahoo! Research,
“ New Sciences for a New Web”

Hector Garcia-Molina, Stanford University Computer Scientist
Moderator for the Panel  

As you might surmise from the names of the speakers and the titles of their presentations, the afternoon was centered on finding and extracting information out of the avalanche of data that we are storing.

Paul Strong talked about the infrastructure that is required to support the massive transaction requirements at eBay and his research into the Semantic Web.

Brewster Kahle who has founded several companies and whose IP is used by 90% of web browsers today talked about his vision and his progress toward providing universal access to all human knowledge through his non profit Internet archive. He is attempting to create the library of Alexandria on the web and make it available to all. After his talk, a person with the typical dot.com mentality, can up to him and asked him what was the value in his archive.org. He could not understand that the preservation of knowledge that would be available for generations to come was a valid goal.

Stewart Butterfield, talked about trends in the ubiquity of capture devices, and the perception that participation in social networks no longer feels weird. Most humans don’t care about the underlying technology of the “cloud”.  Tagging the data becomes important and social networking might be used to refine the tagging of content.

I had to leave before Raghavan’s presentation, but I understand that he talked about searching by intent versus content. Today we search based on key words and popularity of page hits. That may not be a true measure of audience engagement and fulfillment.

The growth in unstructured data is tremendous, and the challenges of storing, retrieving and protecting that data is challenging. New technologies will need to be developed to aggregate and correlate that data in the context of information from the eye of the beholder

The presentations that were given at this InfoLab/Hitachi workshop will be posted at   http://infolab.stanford.edu

Edwin Lim, our country manager for Indonesia, hosted a very successful user conference in Jakarta. Prior to the conference, I was invited to participate on a panel with David Schmeichel of Brocade, and Harry Gumelar of the Tax Authority, who had installed a USP V for their new efiling application. The theme of the panel discussion was Compliance and Data Retention Strategies. The moderator for this panel was, a well know journalist and digital media guru, Deriz Syarief, from Bisnis Indonesia. When I was introduced to Deriz before the panel, he mentioned that he had read my blog during his preparation. Deriz did an excellent job keeping us on track and developing the contributions from each of the participants.

Later as I was catching up with my friend, Sulityawan from Indosat, Deriz joined us and I learned that both of them were bloggers. Sulityawan has his own personal blog which he does as a hobby. He blogs about best practices for storage. (This guy lives and breaths storage). While his blog is posted in Indonesian, I recognized references to the HDS AMS power down feature in his recent post on Green Storage. Deriz has his own blog as well as a corporate blog. He recently blogged about our announcement of the Simple Modular Storage announcement, which was in English. There is a very active blogging community in Indonesia, and many of them know each other through the blogs.

Later in the week I went to Kuala Lumpur, for a similar event which was run by Johnson Khoo, our country manager for Malaysia. This was an interesting time to visit Malaysia since they had just had a national election where the ruling part had lost its 2/3 majority. After the election results were in, their stock market declined by 10% and had to suspend trading for the day. In the midst of all this turmoil, I was not expecting a large turnout from our customers and press, but the opposite was true. Since I had the opportunity to talk to some of the people from the business press, I asked why the election results were so surprising, that the stock market should be so dramatically affected. Apparently, the minority party won through the use of the internet, which ran under the radar of the conventional media.

The Economist has a good write up of this on the internet. Malaysia has a high penetration of internet users, which some estimate at 5 million out of a population of 28 million. Bloggers like Tony Pua, and Jeffrey Ooi were instrumental in raising funding and awareness of the issues, taking their appeal directly to the public. The Economist also noted that there is a secondary internet effect where the content is further disseminated by word of mouth and SMS.

There is no doubt that blogging is a major force not only as a way to raise awareness around arcane technologies like storage, but also as a force to change geo, political, and economic directions. It is a force that must be reckoned with as the elections in Malaysia have demonstrated