Rich Freeman via plug on 12 Dec 2020 11:47:23 -0800 |
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Re: [PLUG] CentOS |
On Sat, Dec 12, 2020 at 2:32 PM Joe Rosato <rosatoj@gmail.com> wrote: > > In short I think Stewardship + Time = ownership. Not said outright, but that is what has been happening here no? Don't allow the free option to gain revenue... > Most of the stuff that makes up RHEL is FOSS (and hence the reason CentOS was a thing). It still is. However, running ANY distro is a lot of work, and my understanding is that RHEL is more about providing an LTS platform that is highly supported. You can't just replicate that by forking a github repo. You basically have to monitor everything RHEL does and then do the same, sans anything proprietary they layer on top. It isn't a ton of work, but it isn't particularly interesting work if you aren't being paid to do it. The business model for commercial FOSS projects is that you pay for the boring stuff. The exciting stuff you can get for free anywhere, because that's what all the kids in college want to work on. The part where it can talk to your ERP or whatever from 2003 and the same binary runs for a decade without any need to modify config files is the boring part. The service manager might have changed 3 times in 20 years, but your shell script that starts/stops everything still runs because of 14 layers of wrappers provided by the distro. Python 2 had an EOL announced a year after you licensed the OS, but the vendor is going to keep it alive with backports for the promised OS EOL, including all the libraries they packaged. There are arguments against embracing this model for sure, but that is the model they sell and you're not going to get it from most volunteer operations. You also get a lot of enterprise-oriented bundling. If you want to buy a commercially-supported framework around something like k8s, Ceph, OpenStack, whatever, chances are you can get it with certifications to work with RHEL. You can get hardware certified to run it. That makes it all a bit more turnkey, though you're going to pay for that convenience. I was looking at the Ceph docs and they have pointers to entire racks of server hardware pre-specified for various use cases that they recommend. That stuff is pretty expensive (though cheaper than most competitors in that space), and you can basically get something that works out of the box that way. Plus if you stay in the lines chances are all the vendors will be testing their updates, and when stuff gets outdated you'll have a defined upgrade path. That is what companies want - something that just works with a clear supported upgrade path, EOLs for budgeting, and so on. -- Rich ___________________________________________________________________________ Philadelphia Linux Users Group -- http://www.phillylinux.org Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug