“The release of Ubuntu 6.06 (Dapper Drake), back in June, brought not only a new desktop system to the Linux world, but also a server system with long-term commercial support. It has one key advantage over similar offerings from Redhat and Novell; the flexibility of the Debian dpkg packaging system.” Read the review here.
All good in theory, but if you’re going to run a production environment and care about your Job at the same time, you’re not going to install some unsupported love-for-all-humans inspired Linux distro, just because it support ReiserFS (which I wouldn’t go near anyway) over RHEL, and has a nice packaging system (Nothing wrong with RPM’s and YUM anyway).
RHEL = Proper commercial support, that’s where they make money. Hell, they even have an office in my city.
RHEL = Proper commerical support from third party hardware vendors is great.
You ain’t going to be apt-getin’ the latest Gtkpod update for your important corporate server, nor are you going to be able to apt-get that binary driver you need for your proliant cluster.
Leave ubuntu for the unoffical webserver, or the desktop linux box.
On the one hand you’re right, on the other you’re wrong. The only real reason RHEL gets the big bucks is because of name recognition. While I personally wouldn’t use Ubuntu on my servers (the problem with their being a LTS version of it, is that the Main repository is the only one fully supported, and there is not that much actual software in Main.)
Debian on the other hand also has corporate paid for support, as do many other distributions besides Red Hat. So here you are very wrong. Also as long as the package is actually made for RHEL then you’ll have no problems and usually they’ll work very well, but in my experience, if you try to go outside of Red Hat’s packages (or even use things from source code) you’re going to break the RPMs. This isn’t just Red Hat that breaks this way, I’ve seen it happen with Suse, Mandrake, etc.
Yum also is no where near as good as apt-get (yet).
the problem with their being a LTS version of it, is that the Main repository is the only one fully supported, and there is not that much actual software in Main.
There surely is more software in Ubuntu’s Main than in RedHat’s official repositories, no?
I trust that the global efforts of all contributers, including RedHat’s, to the Linux kernel and userland will eventualy increase the stability and usability of many/all distributions. Furthermore I wouldnt depend on hardware which *requires* you to run binary modules -how unsure you want to get?-
The point is, to me atleast, that even without the big names we will be able to run most of our services while maintaining our freedom of choice based on the GPL/GNU environment.
Edited 2006-09-08 09:09
How can people mod this up? It’s simply wrong.
Canonical offer proper commercial support and there are dozens of companies offering support for Ubuntu together with Canonical.
http://www.ubuntu.com/support/paid
http://www.ubuntu.com/support/marketplace
Again, I can’t understand how somone can make such an uniformed and simply factually wrong statement in the first place, let alone how people can mod him up.
How can people mod this up? It’s simply wrong. …Again, I can’t understand how somone can make such an uniformed and simply factually wrong statement in the first place, let alone how people can mod him up.
Because he’s probably not talking about the support Canonical offers. He’s probably talking about *commercial* application support, which is pretty minimal right now for Ubuntu. I don’t think Maya, Shake, SAP, and a lot of other high end Linux applications are certified for Ubuntu yet (Oracle as well possibly?).
So, when he’s talking about commercial support, I read that to mean support from ISVs.
There is something called LSB and that is what commercial applications should target (not that great ‘platform’, but it’s a start). Unfortunately, only MySQL and RealPLayer were certified as LSB compliant until now.
I also doubt that RedHat is in a rush to push vendors to adopt LSB. It would hurt their current advantage in high-end (server,workstation) market if high-grade commercial apps start to be officially supported on any LSB compliant distro.
There is something called LSB and that is what commercial applications should target (not that great ‘platform’, but it’s a start). Unfortunately, only MySQL and RealPLayer were certified as LSB compliant until now.
RHEL ships with the LSB platform. LSB is not the issue, nevermind that LSB is extremely limited (relatively speaking) and doesn’t address all of the possible libraries, etc. that an application needs. In addition, most of the time certification has nothing to do with APIs, libraries, or even if the application will run. It has to do with how many resources an ISV has to commit to support each flavour of an operating system.
I also doubt that RedHat is in a rush to push vendors to adopt LSB. It would hurt their current advantage in high-end (server,workstation) market if high-grade commercial apps start to be officially supported on any LSB compliant distro.
Considering they have supported the LSB for a long time…I find that hard to believe.
Edited 2006-09-08 23:11
Ubuntu Server also makes for a killer departmental VMWare Server box.
…nor are you going to be able to apt-get that binary driver you need for your proliant cluster.
I have long since left behind anything that uses a binary driver in Linux and isn’t in the kernel. Why? Because software that is built specifically for Red Hat is of extremely poor quality and gives Linux and open source software a terribly bad name. It certainly isn’t good enough to be used in a productione environment, because then you end up needing the support of Red Hat needlessly ;-).
Running on other distributions is a form of quality check for a piece of software. If it doesn’t pass it, it doesn’t get used.
Edited 2006-09-08 11:35
Where’d you get the idea that RedHat even uses yum?
All good in theory, but if you’re going to run a production environment and care about your Job at the same time, you’re not going to install some unsupported love-for-all-humans inspired Linux distro
Really depends where you work. We actively avoid the pay-for-support Linux distros (RedHat, SuSE, etc) and actively use the community-supported ones (previously RedHat Linux, now Debian, with a few *buntu installs on laptops), as well as FreeBSD.
We prefer to invest our money into training our IT department instead of spending money on support contracts we will never use, or support contracts where all we get are barely trained monkeys at some call-center in Alabama or India.
Hardware support contracts are invaluable and we actively use those. But software support, for us at least, tends to be better on mailing lists, forums, blogs, and man pages than via phone to the vendor.
Obviously, YMMV.
Our philosophy is, if we can’t fix it ourselves, then we won’t use it. And if we can’t get at the source to be able to fix it, then it better be a 10,000x better than the OSS version, or we won’t use it (we have a few commerical apps, like Follett Destiny for centralised library system, eSIS for centralised student information system, a centralised video conference manager — don’t recall the name of it, WebSphere portal) but the rest of our apps are customised in-house, or OSS.
… however, you can subscribe to full commercial support for Ubuntu, from Canonical and its partners. Not just good in theory. 🙂
“… however, you can subscribe to full commercial support for Ubuntu, from Canonical and its partners. Not just good in theory. :-)”
Umm..no you can’t..Canonical is not a true business..they have no income.
Uh, where did you get wild ideas such as this? Canonical is a for-profit business with customers and income.
“Uh, where did you get wild ideas such as this? Canonical is a for-profit business with customers and income.”
Well, according to the website for canonical, they have no product..so..what do they do? They don’t sell support for Linux.
Well, according to the website for canonical, they have no product..so..what do they do? They don’t sell support for Linux.
What?!
What is this page for, then?
http://canonical.com/support
“What?!
What is this page for, then?
” rel=”nofollow”>http://canonical.com/support”
I stand corrected. Last month when I was there they did not offer any services, as I was trying to figure out what they did.
I stand corrected. Last month when I was there they did not offer any services, as I was trying to figure out what they did.
Well, they offered services well before last month, so you must have missed that page.
Perhaps you should investigate the Ubuntu site a little more. 🙂
Server version has 5 year support and desktop supposedly has 3 year support. What I am wondering is how do you know what packages have how many years of support when you are running apt-get?
That is a good question and one that should be answered in some way by Ubuntu/Canonical. As long as the whole process doesn’t get more transparent (and support is being extended outside of the “main” repo), people will think twice before using Ubuntu for corporate servers, I guess.
I wonder how Mandrivas new release would fare on servers in a comparision with Red Hat, Novell and Ubuntu (I only saw Mandriva server reviews twice). Four different and equally great options for servers would be nice imho.
One of the annoying problems with Debian has been its potentially short support lifespan; essentially as long as it takes to get two more releases out. Admittedly this hasn’t been a real problem, to date, but not having firm dates has been an issue in some environments in which I’ve worked.
Debian is currently in the process of adopting a time-based release cycle — new stable release every 18 months. We’ll have to wait and see if they’re able to stick with this plan but at least it looks like Debian’s release schedule is about to become much more predictable.
Another was its perceived lack of commercial support, which often made it very difficult to bring into a corporate environment. While I’ve worked in situations where I had complete authority to use whatever OS I chose, I’ve also been in workplaces where it has been made clear that Debian simply would not be used, due to the lack of a commercial organisation providing security support.
This is not entirely true. Debian offers paid support via third party consultants for corporations that have specialized support needs.
http://www.debian.org/consultants/
I migrated my Nagios server to Ubuntu 6.06 LTS server. Works great, like any other linux I guess.
I’m also in the process of building a Bacula backup server. Right now it’s on Ubuntu as well, but I might just test it on ubuntu and migrate it to CentOS 5 when it comes out early 2007 (wild guess).
I also have this proxy I have to get up and running, and yes, ubuntu will power it.
I’m in my ubuntu phase as you can see. Desktop and servers. But don’t worry, I should get back to Debian proper and CentOS in a while.
How can I trust them after it took two months (not exact sure, but it was long) to find that the root password was in plain text!!!
There is no track record of security, no code auditing, no overflow protection and on goes the list.
I primarily work with RedHat and FreeBSD at work but I have Ubuntu LTS installed on one testbed server. Not bad but it still has some issues that, in my opinion, keep it from being a prime time ready server. Here are just a couple of the most annoying ones.
1. When I say LAMP install, that means the openssh-server package WILL be needed. This appears to be a holdover from being a desktop targeted OS to me.
2. Canonical, please take notes here. If you’re going to chroot a service, please document it somewhere and put the config files in the appropriate place. Prime example is postfix, which is chrroted but has it’s config files in /etc. Took me almost 2 hours to figure out why sasl auth wouldn’t work right. How many people would like OpenBSD’s chrooted apache if they were retarded enough to chroot to /var/www but still put the config files in /etc.
2. Canonical, please take notes here. If you’re going to chroot a service, please document it somewhere and put the config files in the appropriate place. Prime example is postfix, which is chrroted but has it’s config files in /etc. Took me almost 2 hours to figure out why sasl auth wouldn’t work right. How many people would like OpenBSD’s chrooted apache if they were retarded enough to chroot to /var/www but still put the config files in /etc.
You really ought to file a bug report. Posting on OSNews isn’t going to get the Ubuntu developers’ attention. You can file bug reports here:
https://launchpad.net/bugs
You’re absolutely right and that’s exactly what I had already done.
I won’t go near it because Dapper broke my Windows 2000 NTFS partition. It kept reporting the wrong size after I installed ubuntu and resized the partition.