New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Wow. I don't know the quality of the Alma and Rocky support teams, but all the feedback I've ever gotten from large companies that deal with millions of dollars that use Red Hat is that their Support is the best. Complicated kernel stuff getting identified and fixed in two hours with the change getting added to the kernel as they fixed a genuine bug that affects all.
Their customers are the big guys who have the biggest issues and need the best support.
I have no doubt about their skill and expertise. But skill has its limits. Hypothetically speaking, there won't be any bugs in a software that is developed by the best developer.
But that is not the case.
Testing capacity is a different domain altogether. It is QA. It needs capacity (i.e. number of 'testers') as well as expertise and experience.
That is my experience, too. Businesses let alone enterprises don't mind paying big bucks in exchange of support and convenience.
Fair question
It's also a question whether the person writing the answer to the FAQ really meant to say "1:1 binary compatible"
Let's stay away from RedHat and embrace Debian.
It could be a human error, yes. But you probably wouldn't think so if you read their twitter posts.
If it is a marketing statement, they would be doing a huge disservice to the CloudLinux brand, considering some of their latest moves.
whatever it is, they need to clarify that. All I am hearing is a lul. Nobody from Alma came to any hosting community to clarify anything.
Fair enough
I can foresee hard times for AlmaLinux, CloudLinux and... cPanel.
Indeed, "1:1 (binary) compatible" isn't realistic if they don't have access to the RHEL source files
Without the source files, the best that they could achieve (through careful testing) is something like "(very) close 1:1 (binary) compatibility"
No. There's no suggestion that anybody can be perfect. The best developers will always create bugs but you'll identify the best developers by their requirement of QA testing. You'll identify a piss poor developer by their lack of testing appreciation. (I'm looking at you, sickchill Dev).
Yep.
If you make money, you can easily negotiate a license with RH for revenue sharing. It's the nonprofit guys who are in deep.
Let's not dig that hole any further, for good reasons.
Most users reading this thread just need hosting.
The situation is also a bit tricky (if not maddening) in practical terms because the AlmaLinux team have access to the RHEL source packages (minimally, because they have a RHEL subscription), but at the same time, they won't be able to use these packages for AlmaLinux
CloudLinux is a different story because the company could buy access to (i.e., license) the RHEL source packages for their commercial distribution
And how does that benefit RHEL? If not, why would they do so?
If what people are speculating is true, that RHEL blocked their access because they were biting into RHEL's client base, it is wishful thinking but not out of the question.
The end cost of CloudLinix is also a consideration. it won't be any good if it costs as much as cPanel to get RHEL packages.
This is why we need an official statement.
Would not be techically feasible for CloudLinux to switch to Debian/Ubuntu as base OS in the next version?
It is of course feasible, albeit the development itself would be a monumental task as well as testing:
They already have a beta. But I don't follow their developments:
https://docs.cloudlinux.com/ubuntu/introduction/
(Just to be clear, I'm not at all taking RHEL's side here. In fact, I don't think that I fully understand RHEL's decision!)
If CloudLinux will need to pay RHEL extra in order to use RHEL source packages for their (= CloudLinux's) CloudLinux distribution, then RHEL will benefit financially and CloudLinux will lose financially
In order to recuperate their financial loss, CloudLinux would then need to raise their prices for the CloudLinux distribution, but this would be a delicate move (as you say)
Unfortunately, at the moment, the future of free RHEL clones doesn't look so good
Perhaps RHEL will reconsider or at least offer a path forward for free RHEL clones
I think the main take away from all of this is that we all should have put more thought into switching to Debian/Ubuntu based distributions after the CentOS 8 decision. But it seems everyone was quick to embrace AlmaLinux as the heir apparent to CentOS.
You can argue that Debian/Ubuntu didn't take the opportunity to really embrace the void that would be created with the CentOS 8 decision - and that's a valid point. I would argue that Ubuntu probably needs to release a more server specific version of Ubuntu, if for no other reason than to clear it's name as a desktop only distribution. The server specific version could draw heavily from the desktop Ubuntu distro.
After many months of debating, we did decide to go the AlmaLinux path like most everyone else coming from CentOS. But I can't really say that any of these RHEL decisions really comes as a complete shock to me.
Let's stay away from fragmented Linux distros and embrace neatly managed macOS.
macOS needs no license fees.
But the hardware costs 5x.
Now the *BSD world comes pre-fragmented so no confusion about future direction.
I wonder if you can run a server on android.
This thread is about an operating system, not basic web hosting.
Licensing software is lucrative if there's no hardware and support involved. Very lucrative.
https://blog.cloudlinux.com/cloudlinux-os-8-and-9-in-post-redhat-world
CloudLinux plans to release a free version of CloudLinux OS 8 and 9 that will get security updates for it 10 year lifecycle. The OS will be an in-place alternative to AlmaLinux/RockyLinux for customers who are concerned about using AlmaLinux or RockyLinux. The OS will not have all the extra features of CloudLinux OS, such as LVE, CageFS, AccelerateWP, HardenedPHP, etc…
CloudLinux will provide full 10 year lifecycle for CloudLinux OS 8 (May 31, 2029) and CloudLinux OS 9 (May 31, 2032). The support will include all needed bug fixes and security fixes for the subset of packages relevant to the service provider industry. We commit to supporting newer hardware in CloudLinux 9. It might mean that we will use and maintain a non-RedHat kernel in CloudLinux 9.3 and later.
Are people sticking with Alma or Rocky, and if so, which one? I really want to place a bet to stay with either Alma or Rocky, whichever has the biggest chance of staying alive and reliable.
Rocky outlined some next steps, https://rockylinux.org/news/keeping-open-source-open/
Alma says they'll maintain their timely updates despite extra effort, but didn't detail how, https://almalinux.org/blog/our-value-is-our-values/
As rocky is maintained by the original centos author, I will pick Rocky if anywhere want to go from Ubuntu and Debian.
Qeru uses AlmaLinux mainly because Virtualizor does not support Rocky, although I'd much rather use Rocky. My personal servers, web/mail/Mastodon VPS, Tor exit relays, and home server all run Rocky.
AlmaLinux seems to be bigger, and has the backing of CloudLinux. But Rocky could still survive nevertheless.
Or it could die a la Scientific Linux, where CERN thought CentOS was good enough, just for CentOS to die and CERN is forced to adopt AlmaLinux.
It's not like CyanogenMod where we had one definite successor LineageOS, it's more like Mandriva where there are many Mandriva successors: Mageia, OpenMandriva, and PCLinuxOS. It's just that unlike Mandriva clones, RHEL clones are critical OSS projects and far bigger, and RHEL still exists, even if we don't like what it became.
not sure at move to Debian/Ubuntu reaction. doesn't centos stream still support 5 year, just like ubt and Debian does?
Oracle statement!
https://www.oracle.com/news/announcement/blog/keep-linux-open-and-free-2023-07-10/
https://www.redhat.com/en/resources/centos-stream-checklist
"CentOS Stream may seem like a natural choice to replace CentOS Linux, but it is not designed for production use."