Is Your Linux Distro Here For The Long Haul?

Trying to pick the right Linux distro for your small business? Read this before you make another move.When you're kicking the tires on Linux distros, it's easy to get caught up in the technical details. That is true whether you're looking at desktop distros, where software packages make a world of difference, or at server distros, where stability and performance are key concerns.

But what about the organization behind the distro? Can you be sure that the company or the developer community behind a particular distro will be there for the long haul?

Caitlyn Martin, a Linux consultant and O'Reilly blogger, recently raised this issue. Her O'Reilly blog post, and a related post on her personal blog, should be required reading for any company weighing the pros and cons of various Linux distros.

The catalyst for Martin's discussion: A recent uproar over CentOS, a popular business Linux distro. A group of CentOS developers, upset over what they perceived as inadequate leadership, threatened to quit the project en masse unless their concerns were addressed.

The crisis passed, and the CentOS community doesn't appear to be in any immediate danger of a meltdown. But the incident left Martin justifiably concerned about the risks businesses must address when they pick a Linux distro: By shining a light on the CentOS development team I was reminded of something I knew all along but never focused on: CentOS is essentially a small, volunteer project. Much like the many hobbyist desktop Linux distributions CentOS is dependent on a relatively few people. If one key developer leaves the project will suffer. If several leave it would be greatly hurt and might not survive. That is the nature of small projects, not something unique to CentOS. It also, in my newly formed opinion, actually makes CentOS a poor choice for business. Martin points out that stability and reliability are key selling points for Linux as a business computing platform. And, she concludes, "that applies to the organization supporting the code as much as the code itself."

That doesn't necessarily mean that commercial Linux distros (such as Red Hat) or no-cost distros backed by a for-profit service and support organization (such as the Ubuntu/Canonical relationship) are the only games in town for business users.

CentOS, for example, is essentially a free of charge, unbranded version of Red Hat Enterprise Linux. When companies want Red Hat code without paying a subscription fee, it often makes sense to seek out these "downstream" distros.

Many other distros employ a similar approach, and RHEL has spawned an especially large number of unbranded, downstream variants. But companies that use these distros must weigh the benefits of getting an enterprise-quality, very low-cost, Linux distro against the risk that the community supporting it will fall apart.

As Martin notes, for example, another distro based on unbranded Red Hat code, known as White Box Linux, was quite popular -- until about two years ago, when it abruptly dropped off the face of the earth.

The discussion following Martin's post is lively, well-informed, and also well worth reading. Many of the readers who objected to her conclusions, however, got caught up defending CentOS against what they see as a bad rap.

Don't Miss: How To Choose The Best Linux For Your Business

It's true that media coverage of the CentOS affair was often overwrought, and some of it bordered on sleaze. But I think it's wrong to focus too much on CentOS here at the expense of the broader issues Martin has raised.

Here's the bottom line. The Linux business/community ecosystem is amazingly diverse, and with good reason: It allows users to pick the exact mix of software, support, and service they want. As a result, it also allows users to decide how much of the support burden they are willing to shoulder and how much they want a commercial service provider to handle for them.

Community-supported distros, including CentOS, are a vital part of this ecosystem. But companies that consider using no-cost distros must accept the risks that accompany them.

It's a complex issue. In some cases, it's pretty easy to recover if a distro's developer community takes a hike; many downstream Red Hat distros, for example, use similar software packages, and it's a relatively simple matter to switch to another package source for patches and software updates.

In other cases, the only option might be to tear out the old distro and start over. That's not a disaster on a development box, but it could be a costly mess when a production server is involved.

It's also fair to assume that companies like Red Hat, Novell, and even Canonical are a safe bet for companies seeking a Linux distro backed by a stable, well-funded vendor. Then again, many users worry that another distro with big-name backing -- Oracle's Unbreakable Linux -- might suffer as a result of Oracle's acquisition of Sun Microsystems.

And of course, one simply can't evaluate the long-term viability of any community-supported Linux distro based on a particular company's balance sheet. Some community-backed options, such as Debian -- one of the oldest distros in existence and one that boasts an enormous developer community -- are likely to outlive many of the people using it today.

Other cases are more ambiguous and require some digging. How many developers actively work on a particular distro? Has the developer community grown over time? Does the distro serve a particular niche (such as scientists or educators) that helps to guarantee its longevity? Do users complain that a distro's developer community is slow to issue critical security patches or software updates?

And then there's the delicate matter of community politics. Does it seem that a distro's developers are too busy fighting to get any useful work done?

Or is that simply a trick question, since some Linux distro developer communities positively thrive on acrimony?

Business users must consider all of these questions when they decide whether the community -- or the company -- behind a Linux distro is fundamentally sound. I can promise you the answers won't always be easy to find, but a little digging now could save loads of trouble in the future.