chuckd said:
Right on, I agree with you in that Fedora probably won't be great to run on production servers. In fact I'm planning on running RHEL.
Either you're presuming DA will support RHEL, or you're planning on moving to another control panel.
Will I change your mind by pointing out that the price for RHEL is annual, and that before you get it you accept an agreement that they can walk into your office and physically audit you at any time to make sure you own one license for each server your run it on?
I wouldn't mind moving to FreeBSD but I don't have the energy to switch our servers.
I never switch pre-existing servers. I create new ones and every few years I clean out old ones. This year I retired a Sun Cobalt RaQ4; it took a bit over a week, and I only got one customer call, even though just about everything changed, including email logins. It was made easy by the fact that we weren't hosting php or MySQL on that server, so it held only static sites.
I also retired a Plesk system; that was a bit more complex because DA has a different naming convention for MySQL databases and users. But it went without a hitch as well.
In both cases we did it because we didn't trust the underlying operating systems anymore; the RaQ was running a system based on RHL 6.2, and the Plesk server RHL 7.1.
Ahm yes but then how many new releases did Red Hat used to make in a year?
It's not just how many releases but how long the releases were supported. And for fedora, that remains to be seen.
There's no reason why it should be worse with Fedora - you wait out the releases until they're good enough or the current one's been obsolete (and that's generally the fault of the requirements of packages such as PHP & Perl long before RH's EOL).
Or they become obsolete because no one's releasing RPMs to fix compromises; that's the biggest problem. With RH's new one year EOL security becomes a problem long before feature creep. Did the fedora project indicate how long their releases would get security updates?
RHL, on the other hand, is a quality piece of software, opened up to increase the testing base & save Red Hat money. Red Hat have needed to make only simple changes to the way it's developed. They've already made 2 solid testing releases.
As a business user to whom the stability of hosting software is of primary importance, I can't rely on hopes, guesses, and interpolations. I've got to feel comfortable that my software will stay supported, secure, and reliable. Perhaps Fedora will be the best thing since sliced bread. I won't rely on that, though, until it has a history. Especially since FreeBSD is out there, has a history, and is a very stable, supported and secure platform.
Perhaps I feel better about FreeBSD because I owned an ISP that relied on BSD (actually BSDi OS; the pay-for-license version) for years; I have a good bit of experience with BSD, and I'm not afraid to use it.
There are just too many people interested in Red Hat for it to fail to create a large enough community for support to be continued.
I hope you're right.
For example there's already a project called
Fedora Legacy which aims to maintain support for EOL'd Red Hat products.
Thanks for this link; I'll check it out.
Jeff
.