We first failed at cloud because we were closed: IBM

We first failed at cloud because we were closed: IBM

Summary: IBM has claimed that the first generation of cloud services failed horribly because everyone in the industry tried to do it their own way.

TOPICS: Cloud, IBM, Open Source

SINGAPORE--IBM vice-president for SmartCloud Portfolio and Strategy Mohamed Abdula said that the industry made monumental mistakes in trying to get cloud off the ground.

"I'm proud to tell you that at the first generation of cloud, we failed — not just IBM — the whole industry failed."

"We tried to build a cloud environment that is proprietary — all IBM-based, all VMware-based, all Microsoft-based, all whatever your favourite vendor is-based. "

Abdula said that cloud was actually about self-service, resource pooling and importantly, the ability for organisations to deploy apps to multiple environments.

The latter point in particular meant that a proprietary format was never going to work, he said, so the company has since changed its tune and is now building on top of open standards.

"The only way in which you can make a cloud infrastructure work, and be flexible and make it resilient to changes ... is to adopt cloud technologies that is based on standards that allow you the freedom to know that what you invested — your intellectual property on automation, on the application — is protected regardless of the changes in the environment"

Such standards that IBM is betting on include OpenStack and CloudFoundry, as well as other more widely used technologies such as OAuth, Mongo DB, and HTML5.

The progress that open source communities are making however, is not necessarily at the pace that IBM and other vendors would like, according to Abdula.

"Can those standards work by themselves? The answer is no. Open source and open communities take time to mature."

He said that trying to manage groups as large as 3000 members is a difficult task, and the complexities are only compounded by how frequently the sentiment of the group changes.

"One day they all love one thing and other days they love another thing."

Abdula said that IBM sees its role is to build on top of these open standards, even if it means donating its own intellectual property from time to time, and is hoping that more organisations will adopt the same approach.

Michael Lee travelled to IBM InterConnect 2013 in Singapore as a guest of IBM.

Topics: Cloud, IBM, Open Source

Michael Lee

About Michael Lee

A Sydney, Australia-based journalist, Michael Lee covers a gamut of news in the technology space including information security, state Government initiatives, and local startups.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • History repeats and repeats

    Vendor Lock-in is too tempting to pass up. Open Standards are something they put on their datasheet but at the end of the day it's anything but open. Vendors will embrace open standards as the minimum entry level price but they they "enhance" and those enhancements are what lock-in customers to their environment.

    Unix and Linux are prime examples and OpenStack will be no different. Vendors will tout Open but the offering at that level will be severely crippled forcing you to into enhanced offerings which may be based on Open standards but in reality are anything but.

    Cross Cloud compatibility is your safest bet. When the "Cloud" gets blown off course by strong winds you need the flexibility to hop clouds. Whether you can change clouds in a storm will be determined by tool sets with Cross Cloud capabilities and the Implementation Architecture of your platforms and services.

    Don't get locked in. Make sure you can operate cross cloud or switch clouds at a moments notice or even dynamically by completely automated process. Prove it.
  • Sure.

    Apple failed with its iCloud system, or at least the previous three iterations like .mac, MobileMe - have fun searching, I bet you'll find and collect them all!
  • Janis Joplin moment?

    Cannot help but chuckle at this article. For me many of these "closed" environment fail not because they aren't open, but because the culture doesn't allow them to succeed. In the initial instances of cloud I'd certainly suggest that IBM suffered from a challenge of self cannibalization, a lack of competitiveness with amazon, and not fully understanding where the workloads were really coming from.

    For me there are really two styles of cloud... those that look to serve the needs of regulated enterprises, and those that attract developers with green field opportunities. Treading the middle ground is indeed a dangerous place.

    Many would suggest that even the largest enterprises need some of each... secure, global, enterprise styled clouds for more "heritage styled" workloads, and another for the new scale out workloads that will favor PaaS styles, built for east-west styles and software provided resiliency.

    At least this is my experience. To this end, IBM seemed to tread the middle, neither being willing/able to canibalize it's base, nor attractive to the new developers because of their track record.

    I don't know that "open" is the answer for IBM, but will certainly be fun to watch.