Reader Feedback Required: The Infrastructure Theories

32 Comments

With our first conference behind us and the new design almost finished, it is time to focus on the task at hand: writing. I have been wondering about how to tweak the editorial mix to better meet the needs of our community. To that end, one of the topics I want to focus heavily on going forward is Internet infrastructure.

I think we are in a period of flux when it comes to Internet Infrastructure. How long, for example, can Dell or Sun continue to be box makers; when do they actually become computer utilities? Or don’t they? The old rules of the telecom hardware business don’t make sense in a world in which companies like Google and Microsoft spend billions of dollars buying networking gear. Nicholas Carr talks about these ideas in his new book, “The Big Switch.” What it means is a whole new way of thinking that includes new technologies, new startups and new ideas.

The infrastructure focus as I envision it would run along these lines: Internet data centers and related technologies, entertainment networks, client computers and other gadgets. The approach I plan on taking involves looking at new equipment offerings, bandwidth and capacity improvements, new protocols, startups — and how content companies influence it all.

What would be more fun is if you could help me shape the coverage. Let me know what topics you’d like me to write about, as well as if you’re interested in shorter posts or more contextual work, such as the piece on Google using its infrastructure as a competitive advantage. And who are the people you’d be most interested in reading about? The comments are open, and I eagerly await your feedback.

32 Comments

Tim Thigpen

We are following the entry of Microsoft & Google and others you may know about, smaller names, companies who are deploying health care IT and all of its record keeping and financial info, communications and delivery of care to patients.

If ever there was a industry who could really take advantage of the new distribution you describe it is the health care delivery/providers. Remember when wireless phones just sweep over Eastern Europe bypassing an entire “copper based standard we had know . These game changing approaches you talk about are perfect for industries whose margins are thin and their infrastructure fractured or non-existent. Health care is a huge opportunity.

So keep an eye out as we are for how the tussle plays out in health care, and if you know of ways to save money when it comes to IT spending at federal state and research hospitals please share them and cover those stories. The health care industry needs help from a new circle of open source innovative folks who can lead.

sandip gupta

Well, whether it is the cloud, utility, virtualization, SaaS or hosting.

Automation of operations is something that is always ignored in many of these service providers and most of these business models do not go far because they do not consider the operational expenses associated with these models…remember the pre-SaaS (or ASP days)…

It will be great to hear what some of the companies doing to automate the management and monitoring of the infrastructure.

OM…great foot forward to bring attention to this space.

Kishore Jethanandani

Its all about pervasive computing, stupid. Do you for one moment believe that the data for broadband wireless applications is going to move through communication networks. No way. Not even WiMAX. Application management via communication management is possible but data will go through storage area networks. That is where infrastructure comes in.

miggins

Hi Om,

Great idea. I echo most of what others have said, particularly those topics which are challenging business as usual in infra, such as virtualization, grid computing, Google Amazon and the other ‘big guys’ entrance into this space, etc.

Other post fodder:

When will bandwidth prices start to increase, if at all?
When will ad-supported models run their course, if at all?
What are the increased risks of security breaches with so much data stored in the cloud?

Looking forward to it…

Jeremy Penston

Hi Om,

Thanks for asking, it is always a good idea to listen to your what your audience thinks… But, please remember that the reason people come to your site is to hear what you think about something.

You provide editorial, not specifically news. There are competitors out there who do news – they do it well because they provide also provide a flow of gossip to string it all together.

What I come here for is opinion: clear, dry and sometimes even passionate opinion (as we have seen in your recent Facebook posts) on how things are evolving. Just because you don’t do news, doesn’t mean you don’t have to keep up, but you reflect more on the wider issues of a story than some of the other publishers. You should keep this whatever topics you focus on.

As for topics, it’s hard to breakdown because of “the fundamental interconnectedness of things” on the internet. This list is pretty long: applications, users, identity, location, privacy, devices, LAN, CPE, access, exchange, backhaul, authentication, ip addressing, routing, routers, networks, peering, transit, CDN, data centre, servers, power, cooling, environment, recycling…

It’s a multi-dimensional puzzle because of over the top plays so I’d like to see you cover them all, particularly if you can help us join up the dots with your analysis. The challenge will clearly be balancing quality with quantity because you need to keep a decent flow of articles.

Good luck with the evolution,

Jeremy

Pete Steege

Hi Om, great choice. Internet Infrastructure is fertile ground. What I’d like to see you explore:
– Convergence. How will all the content and access clouds come together – voice, media
– Will competitors all chase Google’s model with me-too data centers, or will a shared infrastructure emerge?
– Consumer adoption of online storage
– Security – where will the millions of disk drives go when they retire? Will they tell their secrets?

I enjoy your work. Write on.

Lew

Om,

Great to see you covering this space. After 5 years of you saying to me, “Dude, hosting is boring,” maybe 2008 will change your mind. Of course you have been right, but times are changing. Infrastructure and how IT in general is done will change radically in the coming 5-10 years. I give our POV on our blog, but it will be great to get more in the know commentators covering the topic. A few things of many you might want to consider:

-Should any company own servers?
-The emerging need of carrier class software and tools for cloud computing providers.
-3 big compute centers or many? How many will really emerge?
-Will DCs get overbuilt again as the new boom continues and technology increases the compute power of each foot?
-How do Dell, HP, Sun deal with the concentration of buying by cloud providers?
-What will differentiate the winners and losers? Technology?
-What role will big players like google and MS play vs specialists like Rackspace?

There is no end to topics. Good luck. Look forward to reading.

Nag

Wow! What a great topic and more importantly the way you are approaching. Firstly thanks and secondly
good that you are getting the feedback from real end users. Here are my thoughts..

DATACENTERS/TECHNOLOGY
– How long, for example, can Dell or Sun continue to be box makers ?
This is changing with Vmware coming on board. I was recently at the
” 2007 Fall Security Summit” at Palo Atlo which took place on Dec 5th 2007. It was said that 80% of Data centers are
looking towards Virtualization. If this happens, it will be an end to the Small Boxes [ I would say it is like
going GREEN in one way, as it save lot of Power and Floor Space also. ]

WEB 2.0 / IT / Security: It is interesting to see Web 2.0 companies servicing in IT.
Most of the Web 2.0 are social networking or eCommerce. I have seen very few one of them focus on IT
Ex :
GotStatus : http://www.gotstat.us/

One very important Feature is the amount of data that is being transferred from the Web Servers to Users Browser
Ex :
Digg : Downloaded 1.88 MB and there were 82 HTTP requests ( When i check today Dec 8th. for the Digg home page)
cnn.com 776 kB and 167 request.

I saw some very interesting articles in this scope as to how all the data can be sent to users browser as
one single requests thus saving the bandwidth and time. This is one very interesting area.

One person i would be most interested in reading about is the (Raju Vegesna )
There is some really good work being done in the field of networking( Ex : performance and affordability of 10GbE networks used in storage )
There are some

We all know him from the ZOHO. I wish there is blog for Semiconductor startups also.
Note : Raju Vegesna is the evangelist for Zoho responsible for promoting Zoho applications.

Cheers, Nag

Jim

Om,

The whole infrastructure thing has been hurting my brain for some time. I would greatly appreciate you providing clarity in the area.

Google, Microsoft and others are building data centers of truly historic proportions. 700,000 sq ft can stack a lot of computers. Does anyone inside and outside these companies know what they plan to do with all of these cycles? Is it just about search? Google considers it a strategic advantage? Is that possible?

A second and even more confusing issue is the role of large corporations. Are they going to be customers of this new infrastructure? Are they and can they intrust their data to these centers? Must they continue to have their own separate facilities?

A third area is type of services that will be provided? OS, MySQL, Tomcat, etc… or will they move into the application spaces?

A fourth, is Intel and CISCO. Intel got out of the data center business 4 to 5 years ago but aren’t they more logical data center owners? After all this is their equipment Google and Microsoft are buying to put in these giant centers.

Finally, what is the social impact of all this? How is it going to change what we do and who we are?

Good luck.

Jim

Aaron Brazell

As a sidenote, it’s a bit jarring to be reading a blog for awhile and then encounter a post like this. It’s like, “Whoa! We’re turning left right HERE!” while the driver crazily grabs at the wheel and the car goes up on two wheels careening wildly onto the road to the left.

Isn’t it more productive to make a change in coverage an organic effect of what you do on an everyday basis. In other words, instead of just saying, “We’re going to now go this way!” you just start doing that as a natural extension of your writing and the audience graciously accompanies you inn that turn and participates all along the way.

I’m just saying. ;)

Abigail Johnson

Om,

Congratulations on this decision. I know you know how excited I am about this. So much so that I blogged about it: http://thehighconcept.blogspot.com/.

You guys will do a great job with this initiative. And I hope it starts are trend among other journalists as well. Because it’s not just important to understand and celebrate the latest applications and/or gimmicks, but also the implications of what’s going on underneath them.

Looking forward to continuing to follow your insights.

Abigail

iain verigin
  • Lookin’ forward to your analysis of Internet Infrastructure.
  • I’d really like to hear about the differences between the Internet companies view ( ie Google’s ) of Internet Infrastructure compared with telco versions.
    ** You’ve mentioned the “open” vs “walled garden” approaches before. It would be interesting to know what this means specifically.
    ** At a box level. What is really different about Google’s Ethernet switches and the traditional switches dominated by Cisco.
  • Put another way — What does this “Horizontal Internet Infrastructure” look like?.
    ** The middle is and will continue to be dominated by mature technologies like optical transport and big routers. These can’t change much because of all the cool stuff being built at the end-points.
  • There will be tremendous change at the endpoints.
    ** The data center ( FFIV like stuff ) & all the software that goes with serving up content and the end-user appliance ( like iPhone & Media boxes built around SIGM media chips).

  • I’m sure you’re gonna do a great job on it.

Ajay

I don’t have any specific ideas on topics to cover but I do think gigaom should stick to journalism and stay away from analysis. I haven’t been too impressed with the analysis I’ve seen here, including the google infrastructure piece (you could reproduce google’s infrastructure from 3 years ago at a quarter of the cost because hardware prices drop that fast, where’s the edge there? Rather, the edge is their software and algorithms). Om is a great reporter- I remember when I used to check out Business 2.0 years ago and practically every cover story had his byline- and I like when he adds his skeptical opinion to whatever startup or fad is going around at the moment. However, I think it’s best to have a dollop of Om’s analysis with mostly reporting, rather than doing analysis alone. Any fool with an internet connection can run around and collect reported facts and form an analysis, and many do, but most cannot devote the time for reporting. As a content business, you have the resources to do the reporting that most blogs cannot do and that some people claim blogs do not do. Your competitive advantage is to focus on being the best journalistic source you can be, and while you may not be as big as the WSJ, you can certainly focus on tech much more than they can.

Aditya Kaul

Hi Om,

Here are some wild and wacky ideas for blogs..totally unrelated to your infrastructure related post…however you asked for it and so you can take it or leave it..

  1. The social impact of technology – Internet, mobile, new media. How does all this affect us a human race?

  2. The importance of good ‘design’ in technology. I don’t know if you cover technology from a design perspective

  3. Technology being used to tackle poverty, economic empowerment in places like India, Africa

  4. Have a global view rather than a Bay area centric view. I would like to see more posts coming out of Asia, Europe. E.g. New Media in India is a completely different ball game. NewTeeVee is actually the OldTeeVee out there..if you know what I mean

Kaulout.wordpress.com

Aditya Kaul

Here are some wild and wacky ideas for blog ideas..totally unrelated to your infrastructure related post…however you asked for it and so you can take it or leave it..

  1. The social impact of technology – Internet, mobile, new media. How does all this affect us a human race?

  2. The importance of good ‘design’ in technology. I don’t know if you cover technology from a design perspective

  3. Technology being used to tackle poverty, economic empowerment in places like India, Africa

  4. Have a global view rather than a Bay area centric view. I would like to see more posts coming out of Asia, Europe. E.g. New Media in India is a completely different ball game. NewTeeVee is actually the OldTeeVee out there..if you know what I mean

hobnobLover

Hi Om.

I just wish that you bring back the old design of gigaom. In my opinion, it was much better than this version. The main page looks very cluttered, and it’s hard to pick up a story to read, if you know what I mean.

Content wise: I’m not too sure what to say… It’s good right now!

hobnobLover

Jay Neely

Hi Om,

I’d suggest tackling issues that GigaOm’s prominence can help you cover. I’d like to see interviews, first-looks, and hosted debates between experts on new protocols. The shift from wired to wireless is also something I’d want to see, especially if you can get some officials from telecom companies to tell us when they’re going to stop charging such high data transfer fees, which are holding back the mobile web.

Any smart person who’s a good communicator could provide some kind of context to the topics you’ve mentioned. But you have the resources and connections to provide very special kinds of context, and I hope you’ll do so.

I’ll second Mike’s suggestion of thinking about entrenched enterprise networks, too.

Best,
Jay

Markus Goebel

I would like to read more geeky infrastructure stuff. With a broadband connection you don’t need an oldfashioned telephone, radio, tv, CD collection and newspaper anymore. And everything comes for free. That’s a fact today, although coporate guys always try to tell you that over broadband it’s not reliable and you better stay with your oldfashioned services.

Mobile broadband takes this to another level. And, yes, it works great. With 3G and Wifi I can take all this with me on my phone. That’s the new digital lifestyle I would like to see more on GigaOM. It’s already possible here in Europe with a new infrastructure: cheap 16 MBit DSL and 3G.

Mike

Om,

I think this is a great topic, but how about the (occasional) look at the large, entrenched enterprise networks. Rapidly adopting emerging tech is relatively easy for small companies, or home users, but how do large enterprises take advantage? What technologies help scale an app to say 100K users worldwide? This is the large “Gray” chunk of the internet that isn’t sexy, but accounts for a huge portion of usage.

How can a network that big successfully break away from the Windows/MS Office environment? How can security concerns be effectively balanced against functionality and accessibility? How do these slow-moving behemoths deal with the expectations of younger, more wired employees? How do you use emerging technology to solve the problems that can’t be solved by throwing money at them – COOP, policy based access, and breaking legacy network, application, and data dependancy.

Rahul Lakhani

Om,
I see your vision on how internet infrastructure is changing rapidly and what the future holds. I agree with you in that adding more coverage will generate lots of interest in the reader community. I would love to see your take on what are the main drivers of internet infrastructure are. Is it cost savings or reliability of data in the “cloud”? From a consumer perspective I envision cost is not the main driver but reliability of service is. From the enterprise perspective cost savings would play as important a role as reliability.

Ralph Poole

Hi Om,
Thank you, this is a great topic to explore. I think you need to define the kinds of competitive advantage that each of the players is carving out and how potentially they will win. The investments that Google is making now provide a barrier to entry into the search business as it is currently defined. Microsoft must try to meet Google’s investment to stay in the game. Microsoft may be innovative on the margin, but they are clearly the follower. The only way others will be able to compete is to find a niche or to change the game. I would like to understand what new computing models or business models out there that could threaten Google or Microsoft. Google has demonstrated that they can change their business model rapidly, but can Microsoft? Does either company understand their customers requirements well enough to be nimble when they have such huge investments in infrastructure? Or, will the infrastructure allow them to be more innovative? Will they “stick to their knitting” or will they defuse their energy so widely that they will become vulnerable?

Just a few thoughts on the fly. I am looking forward to the debate.

Ralph

Mahesh R S

Interesting thinking here.

I am a bit surprised though that you are referring to “data center” technologies as Internet infrastructure. Maybe I am being overly restrictive in my understanding, but, for me, Internet infra would typically be the large routers and switches being deployed in the access, edge and core of the network along with the cornucopia of transmission technologies. Given this difference in understanding of a popular term (infrastructure), can I request that you to clarify what you mean by infrastructure in your first post on this topic.

Considering that I enjoy reading your reports, I would definitely like to read long posts. But, I would suggest that you put a summary/abstract at the top of long reports. That way your readers can star interesting reports for later reading and filing in the feed reader.

Regards,
Mahesh

Deepak

Very much like you to explore this, since I am pretty sure you will bring a slightly different perspective than Nick. One of the interesting things here is that while Google is using infrastructure as a competitive advantage, in the majority of cases, the infrastructure is the ultimate commodity (the other Carr book comes to mind here “IT doesn’t matter). How these two worlds balance each other out will be interesting to read.

Venkatesh

Om,

I would be interested in looking at what Amazon is doing with WebServices or SalesForces is doing with its SaS CRM. Dell with its current mindset cannot transition to utilities. At the most they will provide remote back-up facilities but beyond that its a big if. They don’t understand software, period.

If i were re-searching this piece, i would take two case studies.

  1. Startups which are leveraging Amazon like WebServices approach to build their business. Not burning through their initial investment in cap-ex intensive hardware/network investments. Pay as you use and concentrate on innovating in the core business. Leverage internet infrastructure which companies like Amazon are offering.

  2. SMB business that have moved their business to SaleForce IT and thus eliminating complex ERP/CRM systems and thus high IT maintenance costs. It will be interesting to get their perspective on how outsourcing IT infrastructure and core CRM/ERP helps or hurts them in the long run.

I will leave the protocols to another comment. this one is already quite big:).

brian

I want to see offerings in how to do such more economically.

but whatever you cover is great, being a good journalist, not just a reporter on news.

YOu are doing very well!

Geva Perry

Definitely a topic that doesn’t receive enough attention as it should. We’re undergoing a major infrastructure shift — the kind that happens once a decade. One of the least talked about aspects of this is the infrastructure software. Companies such as Google, Amazon, eBay and others have made their scale-out architectures relatively well-known, but there has been little talk about the changes in middleware/application platforms and how they fit in the overall trends in the industry as we are moving to “scale-out”, “cloud computing”. There are sights such as http://www.highscalability.com, who address it, but they tend to go more technically deep. You should be covering the business aspects of it, such as innovative products, whether from commercial vendors such as my own company – http://www.gigaspaces.com – or open source products such as memcached, hadoop, etc. Ther is a lot of innovation and interesting start-ups in this area that do not receive as much coverage.

Kar

A personal bias perhaps but I would like to see the impact of how some of the trends like “cloud computing” affect some of the networking infrastructure vendors. For example, if we move towards to a utility computing environment, the bulk of the spend may move from enterprises to fewer but bigger players. This may affect the bargaining power of the infra vendors negatively. On the other hand, if the needs of the utility computing players are so specialized, will they be better off backward integrating – similar to what Google does with their 10G switches?

Deva Hazarika

Om,

I’m definitely interested in more contextual pieces. I think there are two types of pieces like that, both of which are valuable.

One category is “deeper thought” type pieces where you analyze markets/trends/etc.

The other is simply wrapups with some very light contextual thoughts, but then just a set of links to recent relevant pieces across the GigaOm network.

I’ll respond later w/ some specific infrastructure areas I think would be interesting and important to cover.

rohit

i’d love to see:

  • Apps that are a changing (Web n.0, SaaS, Enterprise Apps) and their impact on infrastructure they need.

  • Gadgets and clients of all the ‘clouds’ we are going to have – from iPhones to virtual desktops/servers

  • Networks & Switching inside the computer, inside the data center, and inside the ‘clouds’ that connect the Apps to the Gadgets/clients

    the emphasis on everything ‘new’ like you say is great – i dont much care about v10.0 release of a cisco switch or juniper core-router.

    if even a minor fraction of what we do everyday within the “Enterprise” moves in to the ‘cloud’, we are in for the mother of all Infrastructure build cycles – much bigger than the 90s… and I am betting that a whole lot more than just a fraction moves in to the cloud.

rohit

Comments are closed.