Cloud Native / Linux / Open Source

Jack Aboutboul: How AlmaLinux Came to Be and Why It Was Needed

6 Sep 2021 6:00am, by

The AlmaLinux Linux distribution came out of nowhere to become one of the first drop-in replacements after Red Hat altered the source of CentOS. Although it’s been almost a year since that fateful move, the dust is still settling and no one distribution has become the top contender to stand as the de-facto replacement. Only recently, Johnny Hughes, one of the real original founders of the CentOS and a current team and board member publically endorsed AlmaLinux.

But AlmaLinux has been certainly making a strong case for itself. It’s a fantastic distribution that is a 1:1 binary replacement for Red Hat Enterprise Linux. I have the pleasure of chatting with Jack Aboutboul, who is the community manager and a board member for AlmaLinux, about this new distribution. Let’s dive in and see what he had to say.

Why did CloudLinux feel the need to create AlmaLinux?

We read Red Hat’s announcement. The community was in turmoil. There was a lot of miscommunication and confusion. Many very smart people made the decision to leave the ecosystem and devote their efforts elsewhere. Despite the announcement, we are still big fans of Red Hat engineering and we still feel that RHEL and its ecosystem are the most stable and reliable platform to build on. We really wanted to make sure that there would always be a free alternative to RHEL — so we started thinking.

We thought hard about why this had happened, what lessons could we take away, and how to prevent it from happening again. I’ve spoken with CentOS maintainers in the past and I remember just how overworked and understaffed they are. It wasn’t a glamorous job by any stretch. They had to be loyal to two houses, their employer and the community. Some of them did most of the CentOS work in their spare time alongside another job and there weren’t many of them to begin with. Ultimately, people would just get tired; it’s a lot of tedious work with no replacement, and no end in sight. That was one factor. Another one, and maybe more importantly, was the reason Red Hat was able to acquire CentOS and get all the trademarks/copyrights/etc. in the first place. That was a consequence of the way the project was structured from the outset. We felt that needed to be addressed as well.

There was also the question of, and really our main concern was, who is going to continue to maintain the distribution. Having other forks is nice, but who will continue to maintain them in three years, five years, and beyond. Who will sponsor them? Who will own that and who will own up to that responsibility.

So at that point, we came to the realization that we were already re-packaging RHEL for 10 years, and we knew that we possess the experience to do it really well, and for a long time. So we started AlmaLinux because we recognized that we were in a unique position to give back to the community on the technical front and to make sure the mistakes of the past with the ownership wouldn’t and couldn’t happen again. It is very easy to be enthusiastic at the beginning of the project. It is much harder to be enthusiastic five to ten years later if no one pays you for that.

What do you believe other enterprise Linux distributions are missing that AlmaLinux delivers?

Our core mission is to provide a 1:1 RHEL rebuild, and that is still our #1 priority. As more and more people join the community, we understand that the future is much, much bigger than that. One of the key things we deliver on is governance and ownership, as mentioned was one of the issues with CentOS. We’ve set up a true 501c6 non-profit with an independent board of directors and communal ownership, with contributors having direct voting capability and a direct voice in project governance. This is something that the RHEL ecosystem always lacked and we can now build a real community around that which can affect real change. It’s a fundamental game-changer.

On the software side of things, Red Hat created an excellent Linux distribution and positioned it for the enterprise. However, the distribution is great for a very large number of other workloads and use cases as well. We now have LiveCD Images for AlmaLinux, AlmaLinux for Raspberry Pi, a whole range of images for containers. The team is working on adding support for more architectures like PowerPC and s390. We’re also talking to multiple people about setting up their own re-spins based on AlmaLinux. There is so much going on and so much we are building and the beauty of it is, we are just getting started.

Is there anything particular on the horizon for AlmaLinux 9?

Yes. We always have to try and stay ahead of the curve. CentOS Stream 9 has been started, and so we’ve started working on AlmaLinux 9. It is a long road ahead, and it will be very interesting to see how it will shape up with the “upstream” first approach.

Will AlmaLinux continue to be a general-purpose distribution, or will there be purpose-built versions?

The core distribution will always remain general-purpose, and we think that is exactly what people are looking for — that stability. There is also always the flexibility to customize it to your needs and we’ve seen people join the community and create purpose-built versions too. Our Live Media images were put together by community members. We’ve had PMs and engineers from companies like Cisco and Samsung engage with us while creating images of AlmaLinux for some of their stuff which is heavily customized. The s390 port comes in from people joining the foundation as well. The same happened with Raspberry Pi, and hopefully, MacBook M1 soon. So in short, yes, we expect other specialized builds and re-spins to follow.

What is the biggest weakness the open-source community faces at the moment?

There are always challenges, and there always will be — it’s part of existence. To that end, the open-source community has two parts to it, the “open source”, technical part and then there is the human/personal element of “community”. There are weaknesses on both fronts and I don’t think we need to be afraid to admit that, the nature of who we are is to acknowledge things and work transparently to make things better.

If we had to single out a few it would be around security and then diversity and cohesion of the community. As far as security, CVEs and vulnerabilities are the basics, but also the number of supply chain attacks that are becoming more prevalent. We also have to highlight the issue with bad actors a-la The University of Minnesota because if you look closely, there are other bad actors out there taking advantage of the community every day, sometimes more obviously than others, and we need to keep our eyes open.

Regarding diversity, we still need to do a lot more to help give under-represented groups achieve a more audible voice and help clear paths to participation through things like education, creating opportunities, etc. and also continuing to build bonds within the existing community and those who are marginalized within it. People often neglect the human element, and when it’s so easy to just fork when you disagree with someone, the higher ground is actually asking yourself “How can I figure out a way to work together with someone I don’t necessarily agree with.”

What was the biggest challenge to bringing AlmaLinux to life?

We worked hard to correct the flaws of CentOS past. The CentOS ownership structure was never fully formalized. It wasn’t a non-profit, and as a result, it was VERY easy for Red Hat to acquire CentOS assets without anyone being able to do anything about it. This is why we were so focused on creating proper governance and ownership, and making sure that no matter what — it will be nearly impossible for someone to acquire AlmaLinux OS trademarks, copyrights, and the rest of the IP.

People may think that it’s straightforward to do that, but it’s actually a lot of work setting up the mechanisms to hand off control and ownership to the community. We put in the time, effort and investment because we really didn’t want a repeat of the CentOS story, neither in three years nor in 15 years down the road. It’s important to note that we enlisted the support of experts in this area, too, with Simon Phipps (former OSI President) joining the AlmaLinux board and advising us.

What is the biggest challenge to getting AlmaLinux to the likes of Azure and Amazon AWS?

Actually, the Amazon Web Services team was great in sponsoring the AlmaLinux OS Foundation from the get-go, as were the Microsoft Azure team. We can’t say enough good things about them both. There is a large amount of engineering effort that goes into building for a cloud platform and there are such diverse workloads, you really need to nail it — and not just the development, the testing too. So the biggest challenge is really time to get everything done and done right. Azure and AWS have been very involved with us since day one and worked with our team to make sure that all their internal tests would pass, and listed us as first-class citizens.

We had a very different experience with Google Cloud and Digital Ocean. We are quite surprised with how they treated the AlmaLinux community. Based on third party verifiable numbers, AlmaLinux is the most popular new RHEL clone, only slightly behind Oracle Linux which has been around much longer. Yet, for some reason, Google Cloud and Digital Ocean are yet to list AlmaLinux in their systems.

We had a much better experience with other providers like Linode and Vultr, two great supporters of the open-source community — they added AlmaLinux to their marketplaces quite quickly.

Do you have any advice for those who will be using AlmaLinux as a base for container and cloud deployments?

Our advice is to lean in! Use what you find and also contribute to making it better. We have multiple container images available and the bulk of the work on them is being done by community members who want to enhance them and make them as awesome and useful as possible — and build more of them. As a container host, the stability you need and the tools you expect are already there, so it’s ready to rock and roll. The landscape is evolving quickly and we’ll also be contributing to upstream projects in the container space and ensuring those make their way down to everyone as well.

What is your one best piece of advice for those new to AlmaLinux?

We have ISOs on over 140 servers around the world, containers, cloud images and it’s also very easy to convert from any EL8 to AlmaLinux with one command. We have a very welcoming and helpful community on our chat, Reddit, forums, and GitHub, hop on over, make yourself comfortable, share your voice, and don’t be afraid to ask for help.

Linux is turning 30. It’s nuts. Take a look at what we’ve managed to accomplish in that time. It’s simply amazing. It’s more than amazing. Linux is everywhere (except the Desktop LOL, but this year is the year, we swear), it powers our computers, servers, our phones, IoT devices, the cloud, the edge — Linux powers our lives — and we still have a long way to go. We’re gonna power the next big medical breakthrough, the next generation of electric, efficient automobiles, the next momentous physics discovery and the exploration of other planets, the rockets to get there, and even the trajectory of humanity itself. There is nothing we can’t achieve if we work together. How would you like to not only be a part of it but to own it as well? This is AlmaLinux. Join us.

The New Stack is a wholly owned subsidiary of Insight Partners. TNS owner Insight Partners is an investor in the following companies: MADE, Prevalent, Real.

A newsletter digest of the week’s most important stories & analyses.