The Linux Foundation has released some survey findings as it hosts its End User Summit. We agree with many of the findings, and discuss our take below. But we also wonder more about which version of Linux these large enterprises are using for their own infrastructure, for cloud computing and for the technologies and services they build on top of Linux. More on that below, too, but first, some thoughts on the survey results.

One of the more interesting findings from the survey pertains to migrations to Linux from Windows. More than 36% of respondents said their Linux deployments in the last two years had been from Windows. More than 31% were moving to Linux from Unix, 13.5% reported no new Linux deployments and most, 66%, said Linux deployments of the last two years have been centered on new applications and services (greenfield deployments).

In terms of Windows migration, we agree there are significant drivers for Linux over Windows in cloud computing, where more than 70% of Linux Foundation respondents cited Linux as their primary cloud platform. We agree that Linux appears to be the preferred route to cloud computing offerings, both public and private (as covered in our special report, Seeding the Clouds). However, we must also acknowledge that the continued, wider availability of Microsoft’s Azure is having and will continue to have an impact on the market and may help Microsoft mitigate the cloud connection to Linux, according to cloud users and mixed-environment shops with which we’ve spoken. Another point to remember here is that Microsoft, which has changed significantly in its approach and strategy with Linux and open source, will likely support other hypervisors and perhaps Linux with Azure as well.

The Linux Foundation survey also highlights continued gains for Linux at the expense of Unix, with 19.8% of respondents indicating a decrease in their use of the OS (compared with 18% decreasing use of Windows and only 1.8% decreasing use of Linux). Those planning on increased use were 76% for Linux, 41% for Windows and 19.5% for Unix. We also wonder whether Oracle’s end of support for OpenSolaris will perpetuate Unix-Linux migration?

We also saw consistency from our own research in the Linux Foundation survey’s coverage of the drivers for and benefits of Linux. We asked more broadly about the factors driving open source software, not just Linux, but the results from both our survey at the end of last year and the recent Linux Foundation survey do match up. While cost savings was cited by our own survey respondents as the top reason for adoption, flexibility was cited as the top advantage after adoption. Similarly, the Linux Foundation survey showed that 67.5% of respondents cited ‘features/technical superiority’ as the top driver for Linux adoption. Next, with 65.4% of LF respondents, was ‘lower total cost of ownership.’ A recurring theme we are hearing in terms of Linux and open source cost savings centers on licensing. Not only do users and customers report cost savings from royalty-free open source software, but they also cite license maintenance as a key source of cost savings. Basically, because software is open source, organizations do not have to worry about convoluted IT audits and keeping track of licensing across physical, virtual, cloud and other resources. Though there still may be some trepidation around open source licensing, it seems that Linux and open source software represent both cost-savings and simplicity in terms of licensing for many users, particularly in cloud computing.

While Linux Foundation does not delve into the version of Linux, we recently asked open source consumers about their Linux choices. We have covered unpaid, community Linux in the enterprise since 2008 and more recently community Linux in the clouds, but we were amazed to hear the response when we asked 286 open source users about their Linux choices. More than 70% (206) of respondents cited ‘unpaid community Linux, such as CentOS or Debian’ as the distribution they use. 12.6% reported use of ‘paid, subscription Linux, such as RHEL or SLES.’ With respondents able to check multiple categories, another 26% said they used a combination of paid and unpaid community Linux and another 5% cited other options. Our survey pool of more than 1,700 open source consumers is made up of about 65% with 50 or fewer employees, 10% with 50 to 100 employees, 7% with 100 to 249 employees, another 9% with 250 to 2,500, and 6% with more than 2,500 employees, but we were interested by the fact that only 16% of our survey pool claimed to be non-paying open source users. We also acknowledge the Linux Foundation survey was aimed at large enterprises and government organizations with $500m or more in annual revenue and 500 or more employees.

Still, we continue to watch community Linux, particularly in cloud computing uses, and have no doubt it is having an impact on the paid, enterprise Linux market, mainly in terms of pricing, support and flexibility.