Live data from Hacker News

June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

redhat.com

151–160 of 228 posts

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#151

This is a very important milestone for us since we have a lot of Centos 7 servers and can't go to RHEL. From my point of view, there are more or less three solutions: a. Go to Debian (or ubuntu server etc); probably the most difficult since everything will be different b. Go to a fresh Almalinux or Rocky 9 installation; also very difficult since there's a lot of stuff to be moved c. Try to use in-place upgrade using…

We briefly looked at it. We decided to just move everything to Ubuntu. It's been a lot of work, and taken a lot of time, but ultimately we feel that it's been worth it. Ubuntu isn't without it's own problems but we're much happier today and feel that we're more future-proof for it.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#152

Earlier quoted context omitted.

If I'm not running a desktop, why would I consider running Ubuntu at all? I feel the same about people offering a kneejerk suggestion of Ubuntu as when I'm told to use Docker without actually knowing anything about my needs. Just because I'm wanting to use Linux does not mean I need Ubuntu. Just because I'm deploying code doesn't mean Docker is the solution. But 99% of the time I talk to people, it's all Docker this…

There is a certain practicality in running the same thing on the server as on your developers' laptops. Ubuntu and Debian are pretty similar, and using Debian gets rid of things like snap, which I'm glad to be rid of - but it adds an extra level of "well it works on my machine" which I'm not glad for. If it's going to be hard to get that special different version of Java to work on the server, it should be hard to ge…

If the headless server has a UI process running that is absolutely wasted resources. Having the same environment for the dev's laptop and the server seems like such a waste of resources. The two things are not the same, so don't try to fit the square peg in the round hole. Whatever these differences are, they should be able to be determined and accounted for in the deployment.

To me, we've gotten a bit lazy and a bit too focused on the wrong things. I feel like this happened when we started running javascript on the server. That was such a shock to the system and all sorts of things had to be done since it just was not the normal way of doing things. Someone decided it'd just be easier to clone the dev's system and deploy that as a server. Like what the actual... I also totally feel the same way about running a Windows Server. Like why on gawds good earth would you ever think that was a good idea?

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#153

This is really the time to make a decision, do you really need support..ehm sorry insurance? Then RedHat/Suse/Oracle if not, choose a trusted community distribution like Debian.

Nobody thinks they need insurance until their house catches fire

Bad comparison, no one will pay for your downtime/leak, but you can blame redhat/suse/oracle for it. Aka you don't loose your job, but the house is still burned down.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#154
post #69

For better or for worse, this has caused a lot of software vendors to drop RHEL/CentOS support. Many have started to offer Ubuntu support in it's place. Most of them didn't even try to support stream. I'm not sure what the process behind that was. CentOS was free, which is why we used it. Sure, our SAP boxes run RHEL but the other stuff doesn't. We started moving a lot of stuff to Ubuntu as well. Now, Ubuntu is by no…

I am always baffled by people using ubuntu because "free". I get the desire if you want the paid stream. EG, an alternative to Redhat or SuSE as a paid stream. But if you're going to use Ubuntu, why put up with the snaps, the commercialism, the app store, when you can just use debian? Every time I hear a reason why, it makes little sense to me. Even though 95% of Ubuntu is Debian, and Ubuntu wouldn't exist without De…

It's supported, and it's known.

Ubuntu = Linux for a lot of people these days, kind of like Red Hat was to us old guys.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#155

Rocky Linux is a fine successor to CentOS and was created by one of the original founders of CentOS, Gregory Kurtzer. https://rockylinux.org/ https://rockylinux.org/about/ If you need enterprise support RHEL tends to be a default choice. If you cannot afford RHEL or do not need enterprise support, Rocky Linux fills the role that CentOS once did.

If you can't justify the cost of RHEL for your use case, I believe Fedora would be the best alternative at this time since its development has not been affected by the changes recently made to CentOS, and it provides a near identical feature set.

There is no planet on which Fedora makes more sense than CentOS Stream as a CentOS replacement.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#156
post #65

Earlier quoted context omitted.

> this has caused a lot of software vendors to drop RHEL/CentOS support Can you list a few? It's good to keep up to date on how this is changing. Thanks!

Ansys for one https://www.ansys.com/content/dam/it-solutions/platform-supp... Veeam doesn't support CentOS stream or 9, there's been a bunch of things here and there. I notice many are moving to Ubuntu. I had to go through all of our vendors last year, and the nice thing about CentOS was almost everyone supported it. Now there's no 'gold standard'.

It says they support RHEL:

"Red Hat Enterprise Linux (RHEL) 7.8, 7.9, 8.1, 8.2, 8.3, 8.4, 8.5, and 8.6 (64-bit)"

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#157
post #150

This is a very important milestone for us since we have a lot of Centos 7 servers and can't go to RHEL. From my point of view, there are more or less three solutions: a. Go to Debian (or ubuntu server etc); probably the most difficult since everything will be different b. Go to a fresh Almalinux or Rocky 9 installation; also very difficult since there's a lot of stuff to be moved c. Try to use in-place upgrade using…

For what it's worth, we're trying to do a lot of our server upgrades with the Elevate scripting to Almalinux, but man can it be a pain. We have a 30% success rate in testing, and even when it goes right it tends to be 2-3 hours of downtime. The alternative though is to rebuild every server on spare hardware and manually migrate data and configurations over. So it's still a better option once we can figure it out. I'm…

That's a bummer! I know there's a few places that have used it on the 50k+ scale, so once the kinks for your environment, it should be great. If there's anything that can be done to improve the script, we'd love to help! Feel free to join the chat and share what's up in ~migrations.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#158
post #91

Earlier quoted context omitted.

> Red Hat had given a lifecycle for CentOS 8 with an EOL in 2029 I thought that neither Red Hat nor CentOS foundation gave a lifecycle date for CentOS 8. Can you help me find that lifecycle statement?

https://web.archive.org/web/20201101131417/https://wiki.cent...

Thanks! Can we see who updated the wiki? If it’s some random person, then I’m not sure it’s quite the smoking gun I’m looking for. If they are someone more closely associated with the project you’d expect to made decisions about this stuff, then that is very useful to know. Thank you!

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#159
post #129
post #69

Earlier quoted context omitted.

I am always baffled by people using ubuntu because "free". I get the desire if you want the paid stream. EG, an alternative to Redhat or SuSE as a paid stream. But if you're going to use Ubuntu, why put up with the snaps, the commercialism, the app store, when you can just use debian? Every time I hear a reason why, it makes little sense to me. Even though 95% of Ubuntu is Debian, and Ubuntu wouldn't exist without De…

There are many reasons, LTS, security kernel patches etc ... Ubuntu is not 95% Debian, as a matter of fact Debian got LTS only recently, before that you were out of luck. I still have bad memories of going to the DC with a USB stick with network driver on them because the Dell server with Debian on it would not ship with non proprietary firmware / drivers.

This is a thing of the past.

From the firmware wiki at debian.org:

"The Debian project took the decision in October 2022 to create a new repository component non-free-firmware, and include its content on installation media for the upcoming Debian 12 release (bookworm) to make things easier for our users. This change was implemented in time for the bookworm alpha 2 release of debian-installer in February 2023, and all d-i releases and daily and weekly images since then all include firmware."

Previously one had to either download the needed firmware separately, or a bootable image containing it so that it was accessible during install. Many users having problems with proprietary hardware didn't know they could just grab the install image with included firmware at https://cdimage.debian.org/cdimage/unofficial/non-free/cd-in... But I blame Debian for this because their site historically has been excellent at hiding stuff.

Re: June 30th, 2024, will bring the End of Life (EOL) of CentOS Linux (2023)

#160
post #76

Earlier quoted context omitted.

Can you define a rug pull?

Promising to support an OS for 10 years then cutting that support to just 2 out of the blue?

The question I’m trying to answer is who promised what, where was that published, and when was it published. If a squirrel promised me 10 years of support, that’s different than a CentOS maintainer making that same promise on the wiki.
Post reply on HN