Blog Post

10 Reasons Enterprises Aren't Ready to Trust the Cloud

Many entrepreneurs today have their heads in the clouds. They’re either outsourcing most of their network infrastructure to a provider such as Amazon Web Services or are building out such infrastructures to capitalize on the incredible momentum around cloud computing. I have no doubt that this is The Next Big Thing in computing, but sometimes I get a little tired of the noise. Cloud computing could become as ubiquitous as personal computing, networked campuses or other big innovations in the way we work, but it’s not there yet. [digg=]

Because as important as cloud computing is for startups and random one-off projects at big companies, it still has a long way to go before it can prove its chops. So let’s turn down the noise level and add a dose of reality. Here are 10 reasons enterprises aren’t ready to trust the cloud. Startups and SMBs should pay attention to this as well.

  1. It’s not secure. We live in an age in which 41 percent of companies employ someone to read their workers’ email. Certain companies and industries have to maintain strict watch on their data at all times, either because they’re regulated by laws such as HIPAA, Gramm-Leach Bliley Act or because they’re super paranoid, which means sending that data outside company firewalls isn’t going to happen.
  2. It can’t be logged. Tied closely to fears of security are fears that putting certain data in the cloud makes it hard to log for compliance purposes. While there are currently some technical ways around this, and undoubtedly startups out there waiting to launch their own products that make it possible to log “conversations” between virtualized servers sitting in the cloud, it’s still early days.
  3. It’s not platform agnostic. Most clouds force participants to rely on a single platform or host only one type of product. Amazon Web Services is built on the LAMP stack, Google Apps Engine locks users into proprietary formats, and Windows lovers out there have GoGrid for supporting computing offered by the ServePath guys. If you need to support multiple platforms, as most enterprises do, then you’re looking at multiple clouds. That can be a nightmare to manage.
  4. Reliability is still an issue. Earlier this year Amazon’s S3 service went down, and while the entire system may not crash, Mosso experiences “rolling brownouts” of some services that can effect users. Even inside an enterprise, data centers or servers go down, but generally the communication around such outages is better and in many cases, fail-over options exist. Amazon is taking steps toward providing (pricey) information and support, but it’s far more comforting to have a company-paid IT guy on which to rely.
  5. Portability isn’t seamless. As all-encompassing as it may seem, the so-called “cloud” is in fact made of up several clouds, and getting your data from one to another isn’t as easy as IT managers would like. This ties to platform issues, which can leave data in a format that few or no other cloud accepts, and also reflects the bandwidth costs associated with moving data from one cloud to another.
  6. It’s not environmentally sustainable. As a recent article in The Economist pointed out, the emergence of cloud computing isn’t as ethereal as is might seem. The computers are still sucking down megawatts of power at an ever-increasing rate, and not all clouds are built to the best energy-efficiency standards. Moving data center operations to the cloud and off corporate balance sheets is kind of like chucking your garbage into a landfill rather than your yard. The problem is still there but you no longer have to look at it. A company still pay for the poor energy efficiency, but if we assume that corporations are going to try to be more accountable with regard to their environmental impact, controlling IT’s energy efficiency is important.
  7. Cloud computing still has to exist on physical servers. As nebulous as cloud computing seems, the data still resides on servers around the world, and the physical location of those servers is important under many nation’s laws. For example, Canada is concerned about its public sector projects being hosted on U.S.-based servers because under the U.S. Patriot Act, it could be accessed by the U.S. government.
  8. The need for speed still reigns at some firms. Putting data in the cloud means accepting the latency inherent in transmitting data across the country and the wait as corporate users ping the cloud and wait for a response. Ways around this problem exist with offline syncing, such as what Microsoft Live Mesh offers, but it’s still a roadblock to wider adoption.
  9. Large companies already have an internal cloud. Many big firms have internal IT shops that act as a cloud to the multiple divisions under the corporate umbrella. Not only do these internal shops have the benefit of being within company firewalls, but they generally work hard — from a cost perspective — to stay competitive with outside cloud resources, making the case for sending computing to the cloud weak.
  10. Bureaucracy will cause the transition to take longer than building replacement housing in New Orleans. Big companies are conservative, and transitions in computing can take years to implement. A good example is the challenge HP faced when trying to consolidate its data center operations. Employees were using over 6,000 applications and many resisted streamlining of any sort. Plus, internal IT managers may fight the outsourcing of their livelihoods to the cloud, using the reasons listed above.

Cloud computing will be big, both in and outside of the enterprise, but being aware of the challenges will help technology providers think of ways around the problems, and let cloud providers know what they’re up against.

95 Responses to “10 Reasons Enterprises Aren't Ready to Trust the Cloud”

  1. Is the issue with cloud computing really around trust? I think that’s the fundamental issue I have with this article.

    All new technologies incur some risk, but so do traditional on-premise technologies. The risk enterprises face by not having a cloud strategy include:
    – Can they innovate fast enough if they build their own architectures?
    – Are they spending unnecessarily on architectures when they should be trying to solve business problems?
    – Are they putting their data at risk within a series of aging legacy systems?

    Cloud computing is another weapon that IT can leverage to serve a growing list of demands by their constituents (who incidentally are more distributed and more web-oriented). It’s not like everything has to be on the cloud, but everything definitely needs to be cloud friendly.

    Derek Cheng

  2. AppNexus solves many of the “Enterprise” problems of the cloud – Security, VLANs, dedicated hardware, and multiple data centers (reliability and geo-diversity). We enable customers to build complex, load balanced, distributed applications across multiple data centers in just a few hours. We have the scalability and ease of use that Amazon EC2 has, but the finished product, a virtual data center, looks and operates like a multi-location enterprise data center operation. We support both LAMP apps and Windows apps.

  3. Stacey..great way to bring this discussion forward.

    People often forget that the compute infrastructure is dependent on the complexity of the apps. If the apps are going to be simple and can be boxed into predefined compute templates, then cloud surely has a lot of room to play. Which is/may be the case for early stage companies and why they are adopting the cloud…or in many cases standard applications from enterprises may land on the cloud as well. This scenario is no different than why simple/template based sites are hosted at large shared hosting environments at godaddy or vs. complex/transaction based hosting infrastructure is hosted at rackspace and/or savvis.

    Very few enterprises have an implementation of the cloud internall because there are not too many tools/technologies that allows them to do it painlessly. Addionally the enterprise IT is focussed on providing technology platforms (a lot more flexibility and custom approach) based on internal customer needs vs. a “cloud”-like managed service (standardized and templatized). The first one is like building a computer in a workshop by assembling various components vs. the second is where you buy a computer itself.

    The adoption of the cloud within an enterprise will happen over time..initially for simple/standard application and eventually running powerful internal clouds with a lot of flexibility and reliability..

  4. Cloud utility is like mass transport. At certain points in history and specific regions / situations, resource constraints made trains / buses ascendant. But the personal car has not gone away, and neither will the standalone PC or server. When gas is expensive, more people take the bus. When storage & processing power (and for that matter, the space and electrical power behind it) is expensive, more people use utility. When storage, processing, and data center space / power are dirt cheap, more people are ok with spending a bit extra on standalone machines.

    But besides cost, there are many reasons why people buy and drive cars – it’s a form of self-expression, a hobby, and a thrill of being in control. The hobby / thrill aspects are relevant to what we call the “server-hugger” – someone who won’t even outsource the data center or won’t put it more than 10 minutes’ drive away without a good reason. And that’s something that has little place in well-run IT. But the parallel to self-expression through cars is competitive advantage through IT – like the customers we have that are paying a premium to put their servers within a few milliseconds of electronic exchanges they’re trading on. Mass-produced utility compute grids won’t be good enough for them any more than a bus will be good enough for a Porsche owner.

  5. Interesting article – let me offer another vendor perspective from ElasticHosts (we provide UK-based virtual servers on our cloud computing infrastructure):

    #1 – Security: Cloud Computing is as secure as any 3rd party data center or hosting provider – and IT outsourcing is very well established and in use by most companies and industries.

    #2 – Logging: Since our customers have full root access to their virtual servers, they can log anything that they wish.

    #3 – Platform Agnostic: A good cloud vendor will provide the ability to run images of any operating system – for instance, the ElasticHosts platform is based on Linux KVM virtualization, and so will run images of any PC operating system and software: Windows, Linux, etc.

    #4 – Reliability: Any hosting service, internal/external, cloud/not, can experience downtime. Cloud providers at least have the option to migrate customers services elsewhere in their infrastructure cloud – for instance, ElasticHosts stores clusters physical servers, using RAID 6 to keep customer data secure, and can bring up customer virtual servers on different physical hardware in case of failure.

    #5 – Portability: See #3 – with platform agnostic clouds, there will be few or no platform portability issues.

    #6 – Environmentally Sustainable: The McKinsey/Uptime report “Revolutionizing Data Center Efficiency” clearly identified virtualization as a key to maintain good levels of utilization and energy efficiency. Cloud vendors are in a unique position to optimize their allocation of virtual to physical servers and hence attain better energy efficiency than traditional hosting.

    #7 – Physical servers: Location of physical servers is indeed important for legal jurisdication (e.g. US Patriot Act National Security Letters and the EU Personal Data Protection directive). Cloud vendors will have to provide clouds physically located in multiple jurisdictions, near our customers – for example, ElasticHosts is one of few cloud vendors with a UK location (or indeed a European location!).

    #8 – Speed: Network speed is indeed critical, but a cloud vendor physically located near the customer (see #7) will be able to provide good latency and bandwidth (e.g. as a UK-based cloud vendor, ElasticHosts can provide 1 Gbps bandwidth and <5ms latency to customers on the UK internet backbone).

    #9 – Internal Clouds: Whilst big firms may have large in-house data centers and IT shops, it would be extremely unusual for these to offer the advanced feature set that today’s 3rd party cloud vendors can provide – e.g. the speed of provisioning, scaling and web-based management.

    #10 – Bureaucracy: I would place big firm bureaucracy as a reason that big firms will use 3rd party cloud vendors – the internal IT shops will not react quickly enough to provide an alternative.

    Richard Davies

  6. Great list, Stacey. I spend a lot of time talking to the enterprise set at events like Interop, and it’s a whole different world from the Web 2.0 startups. It’s easy to forget what “the enterprise” means (and like many of the commenters here, I think these are problems that will eventually be overcome.)

    But these are exactly the kinds of things that prevent, say, a Connecticut-based insurance company, or an HMO in Philadelphia, or a bank in the Southwest (to name three I spoke with recently) from embracing on-demand technology.

    To be sure, cloud providers will offer up portability, or the ability to run a “hybrid” cloud that’s partly in their data centers and partly in someone else’s. But each time the cloud provider caters to “traditional” enterprise needs, they’ll do so at the expense of operating efficiency.

    For example, if a cloud provider agrees to let a customer dictate where their bits are stored, that provider loses the ability to move data around for efficiency and disaster recovery purposes. Every compromise is an inhibitor to a truly on-demand world.

    Imagine that customers got to say which dam their electricity came from, or what kind of coal was burned to make it. Consider how much more complex everything — from sourcing to billing to auditing — would be. That’s analogous to the challenge true clouds face.

    Great, thought-provoking, piece.

  7. interactivechatsystems

    Amazon/AWS is not ‘LAMP-Stack’-centric as asserted in this article. You may run any stack you want to on Amazon and there are plenty of public images out there to support anything from a bare Linux server all the way up to full applications and frameworks in between. This article appears to be mis-guided on many levels, better to do better research in the future lest the entire article’s veracity be undermined.

  8. Sekhar Ravinutala

    What I meant to say was: power consumed in all local servers > power consumed in clouds with same capacity. I.e., the Economist article’s argument doesn’t really apply here because for each server node you bring online in the cloud to add capacity, you’re taking off a bunch of local servers, so there is a net power saving. In fact, that is a business case for cloud (or utility/on-demand) computing that IBM e.g., has been making for years.

  9. Sekhar Ravinutala

    Stacey, your concerns are technically valid I suppose, but then like Ben Franklin asked, “What’s the use of a newborn baby?” It’s not fair to look at an evolving technology and criticize it for not being there yet.


    (a) Your security point is more to do with external hosting in general than with cloud computing specifically
    (b) Logging, platform, reliability, portability, and speed issues are technical hiccups that clouds should overcome pretty easily/quickly
    (c) I don’t get your points on environment friendliness and physical servers – power consumed in all local servers < power consumed in clouds with same capacity, right?

    I do think you have something on resistance by affected folks (e.g., IT) though – I’m sure they’d come up with all kinds of naysayer arguments (including articles like this one! :)), though the huge economic benefits/business case of clouds should overcome those pretty quickly.

  10. Derrick Johnson

    I completely disagree with everything except #10. Your article seems to be written to put FUD (fear, uncertainty, and doubt) into people’s minds. Do you perhaps work in the IT industry? Perhaps you are trying to justify your existence?

    I’m not trying to be harsh, but please… Cloud computing is here to stay and will only get more prevalent. I would argue that internal email goes down a lot more than, say, gmail. Every time my company’s email goes down, I sit and wonder why we bother paying thousands of dollars per year in salaries to people (and thousands of dollars in hardware, electricity, real estate, etc) when we could just plop it on a web-based solution (not gmail, since they read your emails, but there are others out there). Maybe the new mobileme from Apple?

    All companies, regardless of size, should be looking at renting cloud space and downsizing their own IT departments. It’s not going to solve the problem for all applications, but for email and the like, cloud computing is the hands-down winner in terms of cost, performance, and reliability.

  11. Looks like I am late to the party… Great Article Stacy :)

    Here are my thoughts…

    It’s not secure.

    Cloud computing is not an all or nothing game. Enterprises have been sharing data outside the firewall for many years. The “Big Switch” concept is not how enterprises will migrate to the cloud. They will choose parts of their IT to run behind the firewall and others to run outside the firewall. I have talked to quite a few vendors and enterprise customers who are already using the cloud for non sensitive data. The biggest issue with the “Enterprise” and the cloud is that most global 5000 companies don’t want to disclose that they are using a public cloud. As far as compliance and regulation goes, I have talked to at least one huge financial institution that told me (off the record) that they do a cost analysis on the fine vs. the cost of the regulatory implementation. In some cases they opt for the potential fine instead of implementation costs associated with the regulation. Cloud computing costs will have some impact on those type of decisions. In the early 1990’s the Gartner’s of the world were convinced Linux would never play in the enterprise due to security concerns. Also not all cloud computing is public. Vendors such as 3Tera, Cassatt, and IBM provide public cloud infrastructures.

    It’s not platform agnostic.

    Neither is AIX, HP, or Sun. Also, AWS is not strictly built on the LAMP stack it just happens to be one of the most wildly used stacks. There are successful businesses using Java stacks on Amazon. I am not really buying into the “Lock-in” argument on Amazon. There are plenty of vendors today proving Ruby-on-rails EC2 clouds and is an application running on one of those engines considered a lock-in? At the core of EC2 it is just Linux images. Also, how is managing clouds for an enterprise any more difficult than managing AIX vs. Dell server farms. However, I agree, today, Google App Engine is a lock in.

    Reliability is still an issue.

    If you are implying that enterprises that do geographic fail over are more reliable than public clouds than I agree wit you. However, IMO, at the application level it is a wash and entirely based on the design of the application.

    Portability isn’t seamless.

    Here again if the argument is that moving an application from one cloud to another isn’t seamless than I agree. However, I don’t see that being more complicated than moving an application from AIX to Sun or even worse AIX to Windows. If your argument is that getting data from one cloud to another then I say ditto on the aforementioned application design comment.

    It’s not environmentally sustainable.

    That is absolutely not true with private clouds. Take at look at IBM iDataPlex and Cassatt for great examples of managing power requirements. Also, I suspect the cost of running the famous NY Times TIFF2PDF migration would have yielded a much higher electric bill than $240 dollars.

    Cloud computing still has to exist on physical servers.

    I totally agree with you on this one. However, this should just create a larger ecosystem of regionally based cloud vendors. .

    The need for speed still reigns at some firms.
    Versus the need for cost savings? Here again not all applications will be a good fit for enterprise cloud solutions.

    Large companies already have an internal cloud.

    I believe you are stretching the term “cloud” on this one. Very few large IT shops that I have worked with (around 100 a year for the last 10 years) do not have “cloud” in the way of IBM Blue Cloud, 3Tera, or Cassatt. A lot have huge virtualized pools and some autonomic provisioning however, those are the exception and not the rule. Many of the large financial institutions have been running Grids for many years but here again I am not sure I would consider them the same as what are today calling a “cloud”.

    Bureaucracy will cause the transition to take longer than building replacement housing in New Orleans.

    Agreed most large organizations are riddled with bureaucracy, however, there are bleeding edger’s out there. Enterprise IT leaders that see IT infrastructure as a completive advantage will, IMHO, force the use of clouds. I always say, that when the Mad Money guy on MSNBC starts pointing out IT costs line items for his stock picks the clouds will start rolling in.

  12. Stacey,

    A well thought out/reasoned piece.

    Indeed, there is no “Cloud Magic”. So it nice to see some sane journalism on the subject.

    Truth is that same physical constraints & architectural compromises limited third party “Clouds” as Enterprise Environments. By focusing on a small set of simple applications – “Cloud” providers simply avoid some of the difficult design & management issues that Enterprises face.

    The vendor “bandwagon” marketing is sure to damage the credibility of Cloud computing – long before it gets a chance to be useful. Same thing as with the Grid hysteria a couple of years back.

    Richard Nicholson

    Paremus CEO.

  13. “Most clouds force participants to rely on a single platform or host only one type of product. Amazon Web Services is built on the LAMP stack ….”

    Amazon web services push the limits when it comes to keeping their platform open. The above statement of yours gives an impression that Amazon EC2 service somehow restricts their platform to just the LAMP stack. Not true. For eg., you can use Java/J2EE instead of PHP, lighty webserver instead of Apache , Postgres instead of MySql , so on and so forth.

    But i totally agree with you on the Google App Engine. Their platform is pretty restrictive to the developers.

  14. Hi Stacy,

    Interesting article. At Joyent, we are seeing enterprise users starting to adopt the cloud. Major League Baseball runs it’s chat infrastructure on Joyent. LinkedIn runs on Joyent.

    In both cases, these are large companies running important pieces of their infrastructure on using a cloud computing provider.

    Please let me go through your list point by point. Some of your conclusions are based on assumptions about how clouds work that are not necessarily true for all vendors.

    1 – It is possible to deploy securely onto a cloud, and effectively remain behind your firewall. For example, Joyent Accelerators give you root access to a Solaris Zone. You can use this access to lock down all ports and then run VPN server software such as to create a tunnel between your data center and your Joyent Accelerator. At that point, the cloud is behind your firewall.

    2 – Because Joyent Accelerators give you root access, you can log anything you like.

    3 – You are correct that many clouds are not platform agnostic, at least not from an OS level. However, this is not always an issue. A PHP, Rails or Python developer cares about deploying on their preferred development stack, and generally, does not care about what runs beneath them.

    4 – Reliability is possible on the cloud if the cloud vendor is completely transparent about the infrastructure they are providing. It is possible to highly available deployment architectures that leverage hardware loadbalancing and 2N+1 redundancy at both the app and state layers. High availability takes planning and forethought, regardless of whether you deploy within a traditional data center or on a cloud. however, it is achievable in both situations.

    5 – You are right that data portability is an issue with some cloud vendors. Our CTO, Jason Hoffman has an interesting exchange with a representative from Google App Engine about just this issue:

    6 – Yes clouds take power to run. However, the fair thing to do is to compare cloud computing with traditional IT data centers. Joyent’s cloud computing data centers have utilization rates that are over 8 times higher than traditional data centers. For the same amount of computing and cooling resources, we help our customers do more. Happy “Loving Clouds” are green.

    7 – You are right that physical locations are important to clients for security reasons, for disaster recovery and for reasons related to legal jurisdiction. However, it is not correct to assume that all cloud hide the physical locations of their data centers from their clients. At Joyent, we will tell you which data centers are running your infrastructure, and we give you the option to select or move to a new physical location if you prefer.

    8 – You are right that speed is critical for many applications. Even when your cloud infrastructure is plugged directly into the Internet backbone, there are issues related to the speed of light and the number of switches between, say Boston and Sydney. One solution is to pick a cloud vendor that is transparent and gives you the option to pick a data center that is near by.

    9 – We are seeming the opposite trend. We are seeing large companies out sourcing some of their infrastructure to cloud vendors. They are doing so because the TOC of running something on the cloud is significantly less that running it in house.

    10 – Sometimes, legacy bureaucracy actually pushes cloud computing. When someone in the marketing department wants to deploy a new “unknown” Ruby on Rails app on the same infrastructure that is used to run the General Ledger and the Email system, traditional IT managers tend to become concerned. Rather than stifling innovation, we have seen situations where these IT managers actually recommend cloud computing as a lower risk alternative.

    Rod Boothby,
    VP, Platform Evangelism
    Joyent Inc.

  15. Stacey Higginbotham

    Keep the comments coming guys. I wrote this hoping to spur conversation about cloud computing and figure out what issues the technology world needs to think about as it moves toward this model.

    Like many of you I agree that certain apps will remain in house and others will be farmed out. None of these reasons are show stoppers and some, like reliability are bureaucracy, are largely psychological. The economics of large-scale computing are too compelling to keep enterprises from embracing the concept. To say enterprises aren’t ready today doesn’t mean they won’t ever be, which is why these conversations are important.

  16. Ernest Nova

    This article is easily titled Top Ten Myths Why Enterprises Are Not Ready to Trust Cloud Computing

    The cloud service provider is aggregating demand and is likely to be able to hire better security talent. Similarly, a large part of their operating cost will be energy related so it will be in there best interest to lower that cost over time to improve their margins. The scale of an hosted provider will reduce other costs beyond what a typical enterprise could do on their own.

    The latency argument is a fumble too – any large enterprise has many locations and whether their apps are in the cloud or in the corporate HQ does not make much difference to latency for distributed users. Neither environment can fix the speed of light in fiber. In fact, there is a flip argument for cloud computing in the financial services where the traders are clamoring to get their applications as close to the stock exchange computers as possible for latency reasons.

    Not every application should or could be put on the cloud – if the application is proprietary and highly platform dependent then by all means keep it in house. There is no scale benefit there by putting it on the cloud.

    The legal jurisdiction issue on stored data is about the only real unknown – and that too is of importance to certain kind of customers. That does not make cloud computing less relevant to the majority of applications and customers.

    p.s Canada’s concerns are a bit ironic given that many governments in the world are demanding that RIM, a Canadian company, host wireless NOC that process Blackberry emails, inside the respective countries for jurisdiction and law enforcement reasons. (I just note this without commenting on the merits of the argument)

  17. I think the biggest danger is use of these services as paid botnets to do attacks or scrape for emails… already we get tons of unidentified bot accesses from AmazonAWS (*, and no easy way to block without hand blocking ip ranges or slower rdns lookup.

  18. Stacey: I enjoyed the article, but disagree with you on two points:

    4: I agree that cloud computing is not 100% reliable, but I would expect Amazon EC2 and similar cloud services to be much more reliable than the average IT department. That is the core expertise of the cloud host.

    6: I agree with your point here, but I think you go a bit far. No one should be claiming that cloud computing has zero ecological footprint. Rather, one should claim that, thanks to the cloud, you can locate the server where it will make the least impact. Why host a server in NYC, the Bay Area or LA when you can put it in middle of nowhere where it will make less of an impact.

  19. Interesting post and I’m sure you will get a lot of response to it. Some quick comments:

    #1 – Security – I would say Cloud Computing is as secure as with any data center or hosting provider. What the Enterprise is not sure of is the concept of cloud security and determining the parameters that define it.
    #2 – Logging – Technology is driving this movement. Generally stating that it can’t be logged is a bit of a mis-statement. This depends on the provider of the cloud, the engineers administering and developing on it and the 3rd parties developing around it. It will be there sooner than you think.
    #3 – Platform Agnostic – Clarification: GoGrid offers Windows AND Linux so you have a choice. The Cloud Infrastructure providers will offer all flavors. I would say opting for an Infrastructure provider (like GoGrid) ensures that you as the Enterprise or Smaller Company Developer can worry about your development items and not the supporting infrastructure.
    #4 – Reliability – with ANY hosted service (whether your bank, a small website, a SaaS or larger providers), you will experience issues/downtime. Computers will break, infrastructures will not perform correctly, architecture can be bad (e.g., Twitter). Look beyond that to what bolsters that service (e.g., Support, SLAs, etc.). It’s not only the Technology but what support it as well.
    #5 – Portability – Cloud computing is still in its infancy. Telephone number portability took YEARS to come to fruition. Portability will come to the cloud as well over time.
    #6 – Environmentally sustainable – I could be wrong but virtualization and the cloud can help with this. 1 dedicated server is very different than 1 virtualized or cloud server running on a grid or node of servers. It’s better efficiency.
    #7 – Cloud Computing on physical servers – I agree with you here. Don’t really see how it could be done otherwise.
    #8 – Need for speed – to add to this idea, putting things into the cloud is actually faster than going with a traditional dedicated path. Faster setup and deployment. Faster scalability. But yes, if you corporation is on the Right Coast and your cloud infrastructure is on the Left Coast, you will have some (minimal?) latency. But this will changes as more POPs (points of presence) come into existence. There need to be studies benchmarking the difference.
    #9 – Internal Clouds – Yep. There will always be internal clouds, especially for the Enterprise or Gov’t. But there will be (and are) providers of internal clouds. Just depends on the needs.
    #10 – Longer due to Bureaucracy – Agree. But this is true with other technology evolutions (like cloud computing). Enterprises are like cruise ships that take hours/days to make a turn. Right now, early adopters (e.g., startups, web 2.0, etc.) are helping to form the industry. Enterprise will come in a few years (probably starts now with internal projects or skunkwork departments that test the technology and then recommend a further rollout).

    Thanks for your post! It really got me thinking!

    -Michael Sheehan
    Technology Evangelist for GoGrid

  20. jamalystic

    Thanks for the really good dose on cloud computing. With cloud computing being touted as the next ‘big thing’, it ‘s nice to have a piece like this one. But as one reader already commented, smaller businesses will definitely be the ones to easily and quickly adopt cloud computing due to cost. I think cloud computing fits them perfectly. But as one expert pointed out, cloud computing can also be use to larger businesses: Where Cloud Computing Makes Sense (

  21. My belief and experience says that small businesses will adopt the cloud at a much faster rate than medium and larger enterprises, who are much slower moving, have greater security and IT issues and sensitivity, and at much larger switching costs.

  22. My guess is that cloud computing is only the future right now for Web 2.0 type companies – which, despite what many bloggers believe, is only a small portion of the computing world.

    One reason I’m not interested in SaaS and “cloud apps”: who owns the data? At least with the MS monopoly, you can reverse-engineer the file formats (e.g. Open Office & such can read Word & Excel files pretty well – excepting macros). But with all your critical data “in the cloud”, you don’t even know how it’s stored. So what happens when the SaaS vendor goes belly-up? (Likely to happen to most startups) Or gets bought by a bigger fish with a different agenda? Or decides its profits aren’t increasing fast enough, and has to increase revenue from existing customers to get its stock price up? (Likely to happen to customers — like it’s already happened to MS customers, eBay sellers, etc).

    pwb’s comment makes sense to me – run open source apps on commodity hardware, on site, hosted, or both.

  23. Satish Sharma


    Some of your arguments are insane.
    #1. By your arguments small cars are more environment friendly than railroad, that’s the what a small data center to a large data center is — power consumption wise, not to mention the other costs.
    #2. Almost all data centers go down — even one run by companies themselves and it’s probably far more than S3 or google going down for a few hours in 2/3 years.

    #3. Others are mostly bougs.. or for idiots to worry about, except your first item, secure is what most C-levels worry about. Rest you have just filled up with junk .. platform agnostic .. come on .. since the history of computing nothing has been platform agnostic.

    Come to think of it; even god isn’t platform agnostic .. that’s why we have religious wars all the time.

  24. Enterprises may well not be ready to trust the cloud, but I don’t think most of your reasons are significant disadvantages of cloud computing. Of course there are a variety of types of SaaS or cloud computing and the detailed issues vary from one to another. But most of the reasons above also apply to the internal servers of any reasonable sized company and by outsourcing to specialists, then in most cases you will get a better service more cheaply. Probably the main issue remaining to be tackled for many companies is the security issue, though again, because it is their core business and essential for their success, cloud computing companies in many cases will take this more seriously than in-house IT departments.

  25. I still think a lot of the SaaS apps will migrate back to user servers. SaaS is booming right now because it’s cheap it’s cheap to get started, costs are predictable and apps are decent. However I think that virtually all companies should develop a sufficient competence in LAMP-based serving to be able to run many of these services themselves. For example, I’d much rather run my own installation of SugarCRM than continue with Salesforce for many of the reasons cited above.