Since IBM’s announcement that it was kneecapping RHEL clones, there’s been a variety of responses to the move.
If you are not familiar with IBM’s latest developements regarding RedHat Enterprise Linux, here’s a summary:
- For years, CentOS was a 1:1 clone of RHEL. You didn’t get support (or the logo) but otherwise it was a bug-compatible clone. This also enabled vendors such as Oracle to release their own clones, which they self-supported.
- All that came to a halt in 2020 when IBM moved CentOS (renamed CentOS Stream) upstream of RHEL instead of downstream. Now CentOS was a development playground, not suited for production, and it wasn’t a 1:1 clone.
- Several new distros immediately hit the scene (Alma, Rocky) to rebiuld RHEL sources so the world could once again have 1:1 bug-compatible clones.
- Last month, IBM dropped a new bomb, stating that under the terms of the GPL it was only required to release sources to customers to which it had distributed binaries. You want the source? No problem, just buy a license and you can get it in the customer portal. But IBM isn’t selling licenses to cloners.
So now users who thought they had a path forward with a free RHEL clone are in a quandry. Swtiching distros can be painful, and often commercial software packages specify which distros they support, with RHEL being prominent. A popular strategy is to run RHEL proper for production but use a clone for development and testing, which doesn’t work unless your clone is 1:1 bug-compatible.
How the Clones Have Responded
After considering their position, different RHEL cloners have made different responses. Let’s look at how they’re adapting.
Alma Linux: We Will De-clone
Alma was the first of the clones to emerge, on March 1, 2021. It was also one of the first to post a reaction to the IBM changes. Today they posted a new update titled “The Future of Alma Linux is Bright” in which they stated “the AlmaLinux OS Foundation board today has decided to drop the aim to be 1:1 with RHEL. AlmaLinux OS will instead aim to be Application Binary Interface (ABI) compatible.”
This is a de-cloning strategy. Alma will continue to be very similar to RHEL. It may even be better, since the flip side of dropping the 1:1 bug-compatible stance means they’re free to accept bug fixes outside of RHEL’s release cycle. As they put it, “While that means some AlmaLinux OS users may encounter bugs that are not in Red Hat, we may also accept patches for bugs that have not yet been accepted upstream, or shipped downstream.”
Rocky Linux: Come At Me, Bro, and We’ll Make You the Next SCO Group
Perhaps the most interesting approach on this list is the one taken by Rocky Linux. As explained above, IBM is limiting SRPM distribution to those who are customers. Well, PLOT TWIST, if you spin up a RHEL instance in the cloud, you are a RHEL customer. So while you’ve got that AWS VM running, why not download the sources? Then you can shut it down and wait for the next release. I’m sure the Ansible playbook has already been written.
They also have found another loophole: UBI container images which are based on RHEL. As they put it, “Using the UBI image, it is easily possible to obtain Red Hat sources reliably and unencumbered.”
Is this approach viable long-term? Who knows. Here’s something to consider: imagine IBM unleases their army of legal orcs from Isengard and slams Rocky Linux with a lawsuit. On the one hand, even though Rocky got $26m in venture capital funding, that’s a rounding error in IBM’s annual legal budget and the drain on Rocky’s team could cripple the distro, not to mention scaring other clones. On the other hand, can you imagine how that would play out in the press? Big bad IBM suing open source developers – it would be the SCO Linux fight 2.0. Ironically, IBM was one of the Linux defenders in that fight. This time around, I bet Oracle would throw a few coins into the Rocky linux coffers. It’d be ugly.
SUSE: Fork You, IBM, and Welcome to the Future of Semi-Clones (RHEL++)
SUSE Linux announced that they are going to fork RHEL. Talk about being opportunistic.
SUSE already has an enterprise offering. On the one hand, this sort of sounds like they’re telling customers “stop using SUSE Enterprise and use our new RHEL fork instead” but that’s probably misreading the marketing. They’d love you to use SUSE Enterprise, but they acknowledge that there’s now a huge market for RHEL refugees.
They’re dropping $10m into the effort but it seems like a long slog to me. If it was a true 1:1 clone, you have a built-in market. When it’s not a 1:1 clone, you have to build the market for what is essentially a new product. SUSE has a wide userbase and many commercial relationships so it’s not entirely tilting at windmills. I have a feeling this is going to be like Alma: as close to RHEL as possible.
What’s fascinating is that there will likely soon be a “slightly better than RHEL” stream of distros emerging. Fast forward a year and both Alma and SUSE have got their new RHEL semi-clones out there, but these semi-clones (RHEL++ anyone?) will have more up-to-date bug fixes. They’ll probably converge on the same set of patches and fixes, so you’ll have RHEL and then RHEL++. You can almost see the marketing: “we’re not 1:1 bug-compatible. Why do we want all those RHEL bugs?”
Oracle: Sticking Their Tongue Out
Oracle and IBM have been rivals since forever. They both compete in the enterprise space. Back in the 70s, IBM invented the relational database, and then Oracle picked it up and ran with it, leveraging the technology into corporate America’s favorite RDBMS platform. Today, Oracle’s market cap is about 50% bigger than IBM’s, despite the latter getting a 66-year head start.
Oracle launched its RHEL clone in 2006. Unlike the other players on this list, selling Linux is a tiny percentage of their sales, which mostly come from the aforementioned RDBMS as well as their vast application catalog and cloud business. Selling Linux was just a natural fit. Most customers say “I need eBusiness Suite” or “I need Peoplesoft” and then look to see which platforms are supported. Oracle has long supported a diverse array of platforms (everything from proprietary Unices like HP-UX and AIX to Windows) but once they got into building their own engineered systems, why not their own Linux?
Tons of Oracle customers buy their software and run it on RHEL and will continue to do so, and Oracle will continue to support that. But there are at least four paths forward for Oracle Enterprise Linux. The first is that they’ll simply close it down and point customers at RHEL, though I think this is unlikely. The second is that Oracle will strike some deal with IBM to officially license/clone RHEL. The third is that Oracle will adopt the RHEL++ strategy outlined above for OEL. And finally they might officially adopt one of the RHEL++ clones.
Unlike other players, they don’t have to worry if major publishers will certify their software to run on their distro – they are the major publisher.
Oh, and ther is a fifth possibility that they outlined in this cheeky blog post: “…to IBM, here’s a big idea for you. You say that you don’t want to pay all those RHEL developers? Here’s how you can save money: just pull from us. Become a downstream distributor of Oracle Linux.”
What Do You Think?
So that’s my take on the landscape, a review of what the cloners are doing, and what the future might portend. What’s your take? Let us know in the comments below!
Related Posts:
- Hey Providers – Want Some FREE Advertising During the SuperBowl? - November 14, 2024
- Inception Hosting is Closing Its Doors - November 12, 2024
- How Will the 2024 Election Results Affect Tech? - November 11, 2024
Leave a Reply