Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Vreme: 11/15/2011 03:56 AM, Alan McKay piše: Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? THe text bellow in only MY opinion, and I am not the member of the dev team, or have any official capacity except being one of the admins in the CentOS Facebook Group. One of the reasons (as much as I understood) is that initially CentOS team was caught unprepared for the fact that CentOS 6 is not build-able from either CentOS 5 or RHEL 6, or even Fedora's, or even any combination of those distros. In the past you could build CentOS 5 using CentOS/RHEL 5 Beta, something like that, I do not know exact details, but it was easy to build it. 1. When RHEL 6 Beta came out, devs were confronted with hostile building environment with missing versions of packages actually used (they had to file bugs against it and wait for Red Hat to release them while chasing around to possibly find those versions faster. 2. In the past there was not many people training to be on the devs team and existing members are volunteers so they have/had limited free time. It was 6-7 years after any mayor/complex building effort, so even active devs had no mayor problems in that period and they were kind of rusty (I hope devs will not take this against me, it is normal for skills lesser used to require brushing up, I know it on my own example). 3. Infrastructure (hardware) and build environment speed and optimization (in terms of software like mock/smock, binary comparison, etc.) was not up to the task at hand. Even disk space was a stretched to the limit to accommodate all versions, srpms, building environments, ... 4. Way of doing thing CentOS pre-6.x was proved to be inefficient and the gap from upstream releases started to prolong. That is when CentOS devs decided to change policy and do like SL team, and create CR repo so they can publish all completed packages as soon as they are available. Scientific Linux has (at least) 2 paid developers and they started setting up (Koji) building environment (long?) before RHEL 6 Beta was released. That gave them starting advantage. Further more, SL devs decided to push SL 6.0 before 5.7 and 4.9 point releases (contrary to CentOS devs) published in same time frame, so to many on this mailing list it looked like SL devs are overall much faster. Their 5.7 update was (I think) few months behind. Currently, CentOS build system should be in much better shape and we will see how it will do for coming 6.2 point release (already in beta). There is much more relevant info, but this should be the jest an I have work to do. -- Ljubomir Ljubojevic (Love is in the Air) PL Computers Serbia, Europe Google is the Mother, Google is the Father, and traceroute is your trusty Spiderman... StarOS, Mikrotik and CentOS/RHEL/Linux consultant ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Ljubomir Ljubojevic wrote: Currently, CentOS build system should be in much better shape and we will see how it will do for coming 6.2 point release (already in beta). Thanks very much for that. I found your account most interesting and informative. I guess one question that I've never seen raised is if there has ever been a suggestion that Centos and SL should combine, or at least work together? They seem to have exactly the same aim. I wonder why SL was set up, rather than offering to help the CentOS team? I saw statistics - I don't remember where - saying that CentOS had 30% of the Linux market, which I found very surprising, though also satsifying (to me). SL had a tiny share. (I remember now, it was someone complaining that Fedora's share was slipping badly.) I belong to what may be the silent majority who don't really care if CentOS is absolutely up-to-date. (As far as I can see, none of the changes in CentOS-6.1 would make the slightest difference to me. I run CentOS on 3 home servers, and Fedora on my laptops.) I was very struck by the ease with which I upgraded to CentOS-6, compared with the nightmare (now hopefully over) upgrading from Fedora-15 to Fedora-16. It reminded me why I would never run Fedora on a server. To me, the reliability and solidity of CentOS are what I relish, and I'm very grateful to the CentOS team for their work. I don't mind them getting a bit crotchety at times! -- Timothy Murphy e-mail: gayleard /at/ eircom.net tel: +353-86-2336090, +353-1-2842366 s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 01:56 PM, Timothy Murphy wrote: Ljubomir Ljubojevic wrote: Currently, CentOS build system should be in much better shape and we will see how it will do for coming 6.2 point release (already in beta). Thanks very much for that. I found your account most interesting and informative. I guess one question that I've never seen raised is if there has ever been a suggestion that Centos and SL should combine, or at least work together? They seem to have exactly the same aim. I wonder why SL was set up, rather than offering to help the CentOS team? SL does betas and CentOS does not for example. I think the way both projects chose to operate is simply incompatible. I saw statistics - I don't remember where - saying that CentOS had 30% of the Linux market, which I found very surprising, though also satsifying (to me). SL had a tiny share. (I remember now, it was someone complaining that Fedora's share was slipping badly.) Fedora is basically an incubator for new technologies and as such not really an attractive system to install for end-users. If you deal with servers you probably go with CentOS, SL, Debian, etc. and if you want a desktop you probably use Ubuntu. I belong to what may be the silent majority who don't really care if CentOS is absolutely up-to-date. (As far as I can see, none of the changes in CentOS-6.1 would make the slightest difference to me. I run CentOS on 3 home servers, and Fedora on my laptops.) I was very struck by the ease with which I upgraded to CentOS-6, compared with the nightmare (now hopefully over) upgrading from Fedora-15 to Fedora-16. It reminded me why I would never run Fedora on a server. I tend to skip one Fedora release and then do a a plain reinstall and copy my old data I need over. Fedora upgrades always sound rather messy. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Dennis Jacobfeuerborn wrote: Fedora is basically an incubator for new technologies and as such not really an attractive system to install for end-users. If you deal with servers you probably go with CentOS, SL, Debian, etc. and if you want a desktop you probably use Ubuntu. I don't really agree with this. If you are using CentOS on servers it is much easier to use Fedora on laptops, since Fedora is so similar in operation to CentOS. In fact CentOS is more or less identical to an ancient version of Fedora. Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. But isn't everyone today using laptops for everyday use? -- Timothy Murphy e-mail: gayleard /at/ eircom.net tel: +353-86-2336090, +353-1-2842366 s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Dennis Jacobfeuerborn wrote: On 11/15/2011 01:56 PM, Timothy Murphy wrote: Ljubomir Ljubojevic wrote: snip I saw statistics - I don't remember where - saying that CentOS had 30% of the Linux market, which I found very surprising, Wow! though also satsifying (to me). SL had a tiny share. (I remember now, it was someone complaining that Fedora's share was slipping badly.) Because fedora, as has been mentioned here by folks in addition to me, is bleeding edge, not leading edge. There's *NO* *WAY* I'd run it at home, much less at work. snip I was very struck by the ease with which I upgraded to CentOS-6, compared with the nightmare (now hopefully over) upgrading from Fedora-15 to Fedora-16. It reminded me why I would never run Fedora on a server. I tend to skip one Fedora release and then do a a plain reinstall and copy my old data I need over. Fedora upgrades always sound rather messy. The preupgrade is what I've been using the last year, and why I'm now building boxes here with 500M instead of 100M root partitions, figuring that it's what's coming for CentOS, eventually. mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Vreme: 11/15/2011 01:56 PM, Timothy Murphy piše: Ljubomir Ljubojevic wrote: Currently, CentOS build system should be in much better shape and we will see how it will do for coming 6.2 point release (already in beta). Thanks very much for that. I found your account most interesting and informative. I guess one question that I've never seen raised is if there has ever been a suggestion that Centos and SL should combine, or at least work together? They seem to have exactly the same aim. I wonder why SL was set up, rather than offering to help the CentOS team? SL is maintained for Scientists mostly in Fermi Labs and CERN, and it has additional Scientific applications/packages. They are also government funded project, and as such must follow some strict rules. Those are main reasons. There are smaller ones, but even those are enough not to think in the direction of joining projects. snip I run CentOS on 3 home servers, and Fedora on my laptops.) I have setup repository for desktop use of CentOS where I have put many packages (~300 compiled and 45 downloaded from non-repo locations) and in process of solving repo conflicts so major third-part repositories can be the basis for nicely formulated Desktop distro. When I finally have enough time I will finish it and offer entire package to public. I hope it will be soon. -- Ljubomir Ljubojevic (Love is in the Air) PL Computers Serbia, Europe Google is the Mother, Google is the Father, and traceroute is your trusty Spiderman... StarOS, Mikrotik and CentOS/RHEL/Linux consultant ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Vreme: 11/15/2011 03:46 PM, m.r...@5-cent.us piše: The preupgrade is what I've been using the last year, and why I'm now building boxes here with 500M instead of 100M root partitions, figuring that it's what's coming for CentOS, eventually. +1 -- Ljubomir Ljubojevic (Love is in the Air) PL Computers Serbia, Europe Google is the Mother, Google is the Father, and traceroute is your trusty Spiderman... StarOS, Mikrotik and CentOS/RHEL/Linux consultant ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphy gayle...@eircom.net wrote: Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change the drive layout or use removable storage, and almost never change the network connections - and you expect the same programs to run for years. On a desktop, the display is the first priority, ownership of certain devices is expected to magically shift to the user at the console, developers will give up consistent device naming for boot speed, and nobody cares if last year's programs still run with this year's OS. -- Les Mikesell lesmikes...@gmail.com ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Les Mikesell wrote: On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphy gayle...@eircom.net wrote: Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick the drive layout or use removable storage, and almost never change the network connections - and you expect the same programs to run for years. On a desktop, the display is the first priority, ownership of certain devices is expected to magically shift to the user at the console, developers will give up consistent device naming for boot speed, and nobody cares if last year's programs still run with this year's OS. I don't agree with that. Some people do want to keep running what they know, and if the budget's tight mark, trying to find a prboom server for CentOS 5 ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 04:31 PM, Ljubomir Ljubojevic wrote: Vreme: 11/15/2011 03:46 PM, m.r...@5-cent.us piše: The preupgrade is what I've been using the last year, and why I'm now building boxes here with 500M instead of 100M root partitions, figuring that it's what's coming for CentOS, eventually. +1 I doubt that. The issue isn't the technology but the support issues that can arise from updating systems between releases. Red Hat would have to test all kinds of update scenarios and not only between two releases but they'd also have to take into account systems that have been upgraded several times. I'm pretty sure they will stick to the service migration update path they are using now. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 02:56 PM, Timothy Murphy wrote: Dennis Jacobfeuerborn wrote: Fedora is basically an incubator for new technologies and as such not really an attractive system to install for end-users. If you deal with servers you probably go with CentOS, SL, Debian, etc. and if you want a desktop you probably use Ubuntu. I don't really agree with this. If you are using CentOS on servers it is much easier to use Fedora on laptops, since Fedora is so similar in operation to CentOS. In fact CentOS is more or less identical to an ancient version of Fedora. That's why I'm running Fedora too but then I'm not an end-user but an administrator/developer i.e. I actually know how to deal with the intricacies of the system and how to keep my system up-to-date in the absence of a direct upgrade path. Users who don't know much about system management cannot really deal with the complexities that arise from Fedoras fast development progress. Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. But isn't everyone today using laptops for everyday use? Desktop in this context basically means a system with a GUI that's primarily used through an attached monitor and keyboard as opposed to a server that has no GUI installed and is primarily managed through ssh/IPMI. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Dennis Jacobfeuerborn wrote: On 11/15/2011 04:31 PM, Ljubomir Ljubojevic wrote: Vreme: 11/15/2011 03:46 PM, m.r...@5-cent.us pie: The preupgrade is what I've been using the last year, and why I'm now building boxes here with 500M instead of 100M root partitions, figuring that it's what's coming for CentOS, eventually. +1 I doubt that. The issue isn't the technology but the support issues that can arise from updating systems between releases. Red Hat would have to test all kinds of update scenarios and not only between two releases but they'd also have to take into account systems that have been upgraded several times. I'm pretty sure they will stick to the service migration update path they are using now. preupgrade is only for migration for full releases, and does sorta kinda work It's been in fedora a year or so; I'm *not* looking forward to it hitting RHEL, and so CentOS, but I'm figuring it will, in another year or two. mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 05:23 PM, m.r...@5-cent.us wrote: Les Mikesell wrote: On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphygayle...@eircom.net wrote: Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick Supermicro boards come with IPMI on-board these days so you can do all that work that you previously did standing next to the server from the confines of your cozy home. This is even more useful when you server is sitting in a rack in a cold, noisy, dry collocation facility. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 10:23 AM, m.r...@5-cent.us wrote: 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick You only need that for installs or if you've done something wrong. And then it isn't really a 'display'/GUI as much as a text based tty emulator. the drive layout or use removable storage, and almost never change the network connections - and you expect the same programs to run for years. On a desktop, the display is the first priority, ownership of certain devices is expected to magically shift to the user at the console, developers will give up consistent device naming for boot speed, and nobody cares if last year's programs still run with this year's OS. I don't agree with that. Some people do want to keep running what they know, and if the budget's tight Then you probably don't run Fedora - the 'desktop' oriented distribution, or care much for the non-backwards compatible changes that went from there to RHEL. -- Les Mikesell lesmikes...@gmail.com ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Dennis Jacobfeuerborn wrote: On 11/15/2011 05:23 PM, m.r...@5-cent.us wrote: Les Mikesell wrote: On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphygayle...@eircom.net wrote: Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick Supermicro boards come with IPMI on-board these days so you can do all that work that you previously did standing next to the server from the confines of your cozy home. This is even more useful when you server is sitting in a rack in a cold, noisy, dry collocation facility. Um, reality check time: what colo? I've got two server rooms, er, computer labs, and a very small one. In the two, we've got maybe 150 or more servers. We don't have them all wired with IPMI. In fact, we don't have any of them cabled that way. Lessee, wouldn't that be an extra port for each server? Or a few servers with their own switches, and all those servers cabled? That's a lot of work for the three of us, *and* there are plenty of times when no, IPMI either a) doesn't work, or b) you have to physically powercycle the damn thing. Or the one that I have to run down to and hit f1 so it'll finish posting. Or be there because I forgot to tell it fastboot before I rebooted it (or it rebooted), and I have to powercycle it, because, as a production box, we can't wait four or six hours for the fsck to complete. (Don't get me started on *that* state of affairs.) mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 05:40 PM, m.r...@5-cent.us wrote: Dennis Jacobfeuerborn wrote: On 11/15/2011 04:31 PM, Ljubomir Ljubojevic wrote: Vreme: 11/15/2011 03:46 PM, m.r...@5-cent.us piše: The preupgrade is what I've been using the last year, and why I'm now building boxes here with 500M instead of 100M root partitions, figuring that it's what's coming for CentOS, eventually. +1 I doubt that. The issue isn't the technology but the support issues that can arise from updating systems between releases. Red Hat would have to test all kinds of update scenarios and not only between two releases but they'd also have to take into account systems that have been upgraded several times. I'm pretty sure they will stick to the service migration update path they are using now. preupgrade is only for migration for full releases, and does sorta kinda work It's been in fedora a year or so; I'm *not* looking forward to it hitting RHEL, and so CentOS, but I'm figuring it will, in another year or two. It might be available as a package but I doubt it will be officially supported by RHEL. sorta kinda isn't good enough for an enterprise OS. If business customers begin hosing their systems with these upgrades then Red Hat will be in quite a bit of trouble. Sure upgrading from a sysv init based system to systemd init based system might work well for your LAMP system but what will it do to proprietary clunky software that is running out there? Will your complex Oracle DB setup actually survive that upgrade? Right now customers have to upgrade by creating new installs that they can test independently of their running infrastructure which makes them ultimately responsible for the upgrade (migration really) process. With an upgrade path between major versions Red Hat will become responsible for that and I'm not sure they are willing to bear that burden for all the possible various installations out there. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 10:57 AM, m.r...@5-cent.us wrote: I don't agree with that. Some people do want to keep running what they know, and if the budget's tight Then you probably don't run Fedora - the 'desktop' oriented distribution, or care much for the non-backwards compatible changes that went from there to RHEL. Hell, no, I don't run fedora. I've got three or four users, and my manager on one of his systems, who do. I *LOATHE* it, with all the grief upgrades have given me. And, correspondingly, you probably don't really run any 'desktop' applications that are visual or audio/video oriented. There are reasons for that side of the coin, but they don't mesh very well with server use and remind me of the early days of Windows. -- Les Mikesell lesmikes...@gmail.com ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 05:55 PM, m.r...@5-cent.us wrote: Dennis Jacobfeuerborn wrote: On 11/15/2011 05:23 PM, m.r...@5-cent.us wrote: Les Mikesell wrote: On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphygayle...@eircom.net wrote: Incidentally, I don't really understand what is meant by the term desktop nowadays. I always think of it as a contrast to laptop. 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick Supermicro boards come with IPMI on-board these days so you can do all that work that you previously did standing next to the server from the confines of your cozy home. This is even more useful when you server is sitting in a rack in a cold, noisy, dry collocation facility. Um, reality check time: what colo? I've got two server rooms, er, computer labs, and a very small one. In the two, we've got maybe 150 or more servers. We don't have them all wired with IPMI. In fact, we don't have any of them cabled that way. Lessee, wouldn't that be an extra port for each server? Or a few servers with their own switches, and all those servers cabled? No, you can share the interface so you don't need any extra cables/ports at all. That's a lot of work for the three of us, *and* there are plenty of times when no, IPMI either a) doesn't work, or b) you have to physically powercycle the damn thing. Or the one that I have to run down You can physically power cycle the system with IPMI. to and hitf1 so it'll finish posting. Or be there because I forgot to tell it fastboot before I rebooted it (or it rebooted), and I have to powercycle it, because, as a production box, we can't wait four or six hours for the fsck to complete. (Don't get me started on *that* state of affairs.) You can hit f1 using the IPMI console. You can also modify the BIOS settings. The IPMI controller is a completely separate system. You can physically shut down the computer and still connect to the IPMI subsystem/web interface and power it back on remotely. Obviously if you don't have IPMI on some systems or cannot use it for other reasons then that's tragic but inevitable. All I'm saying is that for new system you should strongly consider it. Back in the days you actually needed to buy an additional card for this but as I said on Supermicro boards/systems you now get this on-board and it simplifies administration greatly. Just a few days ago I had to re-install a system and in the process change the SATA settings from IDE to AHCI in the bios. In the past I had to go to the server to do this. Together with the managed switches I can completely revamp the entire infrastructure if I wanted to and wouldn't even have to leave my home to do it. Regards, Dennis ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, 15 Nov 2011, Dennis Jacobfeuerborn wrote: Obviously if you don't have IPMI on some systems or cannot use it for other reasons then that's tragic but inevitable. All I'm saying is that for new system you should strongly consider it. Back in the days you actually needed to buy an additional card for this but as I said on Supermicro boards/systems you now get this on-board and it simplifies administration greatly. Just a few days ago I had to re-install a system and in the process change the SATA settings from IDE to AHCI in the bios. In the past I had to go to the server to do this. Together with the managed switches I can completely revamp the entire infrastructure if I wanted to and wouldn't even have to leave my home to do it. Yep, it works really nicely in small HPC machines, where it completely replaces the managed PDUs we'd previously used, and costs you no extra cabling. In my case these are all Dell machines (IPMI's been standard for years on Dell servers). When the IPMI controller can be configured to use DHCP with a poke or two of the buttons on the front of the machine (if it's not been preconfigured) it's really quite quick to rack up and configure machines. jh ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Les Mikesell wrote: On Tue, Nov 15, 2011 at 10:57 AM, m.r...@5-cent.us wrote: I don't agree with that. Some people do want to keep running what they know, and if the budget's tight Then you probably don't run Fedora - the 'desktop' oriented distribution, or care much for the non-backwards compatible changes that went from there to RHEL. Hell, no, I don't run fedora. I've got three or four users, and my manager on one of his systems, who do. I *LOATHE* it, with all the grief upgrades have given me. And, correspondingly, you probably don't really run any 'desktop' applications that are visual or audio/video oriented. There are reasons for that side of the coin, but they don't mesh very well with server use and remind me of the early days of Windows. Um, users here run eclipse, among many other things. At home, I run mplayer, realplayer, browser, gwenview... what audio/video apps were you thinking of? mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Vreme: 11/15/2011 05:58 PM, Dennis Jacobfeuerborn piše: With an upgrade path between major versions Red Hat will become responsible for that and I'm not sure they are willing to bear that burden for all the possible various installations out there. I do not think they will, but 500MB boot partitions I create -- Ljubomir Ljubojevic (Love is in the Air) PL Computers Serbia, Europe Google is the Mother, Google is the Father, and traceroute is your trusty Spiderman... StarOS, Mikrotik and CentOS/RHEL/Linux consultant ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Dennis Jacobfeuerborn wrote: On 11/15/2011 05:55 PM, m.r...@5-cent.us wrote: Dennis Jacobfeuerborn wrote: On 11/15/2011 05:23 PM, m.r...@5-cent.us wrote: Les Mikesell wrote: On Tue, Nov 15, 2011 at 7:56 AM, Timothy Murphygayle...@eircom.net wrote: 'Desktop' is in contrast to 'server'. On a server, you only reboot to load a new kernel and you never use the console display, rarely change Oh, I dunno - it's not infrequently that I have to plug in a monitor-on-a-stick Supermicro boards come with IPMI on-board these days so you can do all snip I understand you love your Supermicro boards. Fine. There's no way we're going to replace everything, which is what you seem to be suggesting. I would also need more ports, to plug in the IPMI interfaces on the boxes we have. Look (and, officially, I am speaking for myself, not my employer nor the US federal government), this is a US gov't agency. Why don't you call your Congresscritter and Senator, and tell them you personally want to donate the money to replace everything we have that doesn't have IPMI, and pay for the time install and cable it all up? That would be *great*... of course, some of our latest servers have 48 cores, and we just got some 64 core servers, so it might cost you a pretty penny Oh, yes, and then there's the official requirement that I be here during business hours. mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Greetings, On Tue, Nov 15, 2011 at 11:27 PM, m.r...@5-cent.us wrote: Why don't you call your Congresscritter and Senator, and tell them you personally want to donate the money to replace everything we have that doesn't have IPMI, and pay for the time install and cable it all up? That would be *great*... of course, some of our latest servers have 48 cores, and we just got some 64 core servers, so it might cost you a pretty penny Oh, yes, and then there's the official requirement that I be here during business hours. I have to apolgise here Dear Mark. Your country has placed so may restrictions on exports that even India (a de-facto nuclear power) does not have the privilege of having such hardware. Even Redhat is afraid. That said, I have had situation in IDC's where I have managed about 45 Rack servers and 10 Blades -- all (then) Sun X and V servers from ground (tile floor up) for an App that is designed to tun on only those servers. I know the enormous difficulties the customer faced, and I had to help them justify. So let us just chill. -- Regards, Rajagopal ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Make centos a new distro and forget about rh 2011/11/14 Alan McKay alan.mc...@gmail.com These seems to me to be the first message in the series and provides a really good summary of the changes at Red Hat which seem to be making life a lot more difficult for CentOS. Just figured I'd pull it out of that thread and change the subject line. Below Johnny's email I've copied another from the original thread, written by Lamar Owen, which gives some good explanation on how Red Hat is able to get away with this. Basically from what I gather, while Red Hat cannot restrict access to sources, they can restrict access to binaries. And since CentOS has a goal of binary compatibility with upstream, they are essentially left trying to hit an unknown target. But (now I'm stretching my limited knowledge even further) Scientific does not have this restriction since they are less concerned about exact binary compat. On Fri, Oct 21, 2011 at 12:54 PM, Johnny Hughes joh...@centos.org wrote: On 10/21/2011 10:01 AM, Les Mikesell wrote: On Fri, Oct 21, 2011 at 9:51 AM, Nicolas Thierry-Mieg nicolas.thierry-m...@imag.fr wrote: Johnny, chill. I don't blame him for being confused. Up until right now, you updated to a point release, then, over the weeks and months, there were updates. All of a sudden, there are *no* updates for the 6.0 point release, which is a major change in what everyone expected, based on history. this is the way it has always been: once upstream releases x.y+1 , there are no more updates to x.y (in upstream and therefore also in centos), until centos releases x.y+1 . Yes, but that used to be transparent, because the centos x.y+1 release happened quickly so it didn't matter that the update repo was held back until an iso build was done. Yes, and NOW the release process is MUCH harder. Red Hat used to have an AS release that contained everything ... we build that and we get everything. Nice and simple. Build all the packages, look at it against the AS iso set ... done. Two weeks was about as long as it took. Now, for version 6, they have: Red Hat Enterprise Linux Server (v. 6) Red Hat Enterprise Linux Workstation (v. 6) Red Hat Enterprise Linux Desktop (v. 6) Red Hat Enterprise Linux HPC Node (v. 6) Red Hat Enterprise Linux Workstation FasTrack (v. 6) Red Hat Enterprise Linux Server FasTrack (v. 6) Red Hat Enterprise Linux Desktop FasTrack (v. 6) Red Hat Enterprise Linux Scalable File System (v. 6) Red Hat Enterprise Linux Resilient Storage (v. 6) Red Hat Enterprise Linux Load Balancer (v. 6) Red Hat Enterprise Linux HPC Node FasTrack (v. 6) Red Hat Enterprise Linux High Performance Network (v. 6) Red Hat Enterprise Virtualization They have the same install groups with different packages based on the above groupings, so we have to do some kind of custom generation of the comps files to things work. They have created an optional channel in several of those groupings that is only accessible via RHN and they do not put those RPMS on any ISOs ... and they have completely changed their Authorized Use Policy so that we can NOT login to RHN and use anything that is not on a public FTP server or on an ISO set ... effectively cutting us off from the ability to check anything on the optional channel. Now we have to engineer a compilation of all those groupings, we have to figure out what parts of the optional channels go at the point release and which ones do not (the ones that are upgrades). Sometimes the only way to tell is when something does not build correctly and you have reverse an optional package to a previous version for the build, etc. We have to use anaconda to build our ISOs and upstream is using something else to build theirs .. so anaconda NEVER works anymore out of the box. We get ISOs (or usb images) that do not work and have to basically redesign anaconda. We can't look at upstream build logs, we can't get all the binary RPMs for testing and be within the Terms of Service. And with the new release, it seems that they have purposely broken the rpmmacros, and do not care to fix it: https://bugzilla.redhat.com/show_bug.cgi?id=743229 So, trust me, it is MUCH more complicated now than it was with previous releases to build. With the 5.7 release, there were several SRPMS that did not make it to the public FTP server without much prompting from us. And with the Authorized Use Policy, I can not just go to RHN and grab that SRPM and use it. If it is not public, we can no longer release it. So, the short answer is, it now takes longer. Thanks, Johnny Hughes Lamar Owen lo...@pari.edu via centos.org Oct 28 to CentOS On Friday, October 21, 2011 02:22:26 PM Les Mikesell wrote: Which is explicitly imposing additional restrictions. Which is explicitly prohibited in section 6. I don't see any exceptions relating
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 11:52 AM, m.r...@5-cent.us wrote: Hell, no, I don't run fedora. I've got three or four users, and my manager on one of his systems, who do. I *LOATHE* it, with all the grief upgrades have given me. And, correspondingly, you probably don't really run any 'desktop' applications that are visual or audio/video oriented. There are reasons for that side of the coin, but they don't mesh very well with server use and remind me of the early days of Windows. Um, users here run eclipse, among many other things. At home, I run mplayer, realplayer, browser, gwenview... what audio/video apps were you thinking of? Vlc is probably the best of the bunch. But most of my laptop video-viewing is Netflix or from a Slingbox so I run Windows on the default-boot partition and run linux under VMware player or more often just connect to a server via NX/freenx for work. -- Les Mikesell lesmikes...@gmail.com ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Nov 14, 2011, at 7:56 PM, Alan McKay wrote: Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? I got the impression that the reason owes to the fact that Scientific Linux is using a koji build server and had it up and running perhaps even before the 6.0 release. http://lwn.net/Articles/446556/ But in truth, don't trust what the non-invested people might speculate to be the reasons, the real answers can only come from the developers themselves. Craig ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 2:36 PM, Craig White craig.wh...@ttiltd.com wrote: On Nov 14, 2011, at 7:56 PM, Alan McKay wrote: Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? I got the impression that the reason owes to the fact that Scientific Linux is using a koji build server and had it up and running perhaps even before the 6.0 release. http://lwn.net/Articles/446556/ But in truth, don't trust what the non-invested people might speculate to be the reasons, the real answers can only come from the developers themselves. And note that one of the SL developers has taken a position at Red Hat, so things might be different in the future http://scientificlinuxforum.org/index.php?showtopic=897 -- Les ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Rajagopal Swaminathan wrote: Greetings, On Tue, Nov 15, 2011 at 11:27 PM, m.r...@5-cent.us wrote: Why don't you call your Congresscritter and Senator, and tell them you personally want to donate the money to replace everything we have that doesn't have IPMI, and pay for the time install and cable it all up? That would be *great*... of course, some of our latest servers have 48 cores, and we just got some 64 core servers, so it might cost you a pretty penny Oh, yes, and then there's the official requirement that I be here during business hours. I have to apolgise here Dear Mark. Your country has placed so may restrictions on exports that even India (a de-facto nuclear power) does not have the privilege of having such hardware. Even Redhat is afraid. Oh? That level hardware is a no-no? I admit, they *are* brand new, and given the very serious scientific computing we do here, they're *needed* (some folks' jobs run, on a cluster like the above, 2, 3, 4 *days*. Then there was the guy about a couple years ago, who asked me to hold off rebooting his home directory server until his job finished. Two *weeks* later, I got to reboot That said, I have had situation in IDC's where I have managed about 45 Rack servers and 10 Blades -- all (then) Sun X and V servers from ground (tile floor up) for an App that is designed to tun on only those servers. I know the enormous difficulties the customer faced, and I had to help them justify. Sun? Debacle, er, Oracle? Oh, *Ghu*, I'm *so* sorry. It took me a month to get one server repaired, getting an FE out, the beginning of this year. (Meanwhile, Dell's had two SE's over the course of three days in the last *week*). If my manager, the other admin, or I have anything to do with it, we *ain't* buying more Sun/Oracle. So let us just chill. Actually no, we turned off the a/c in the room I spent, um, about 7 hours in with the FE between yesterday and today g mark ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/15/2011 06:56 AM, Timothy Murphy wrote: Ljubomir Ljubojevic wrote: Currently, CentOS build system should be in much better shape and we will see how it will do for coming 6.2 point release (already in beta). Thanks very much for that. I found your account most interesting and informative. I guess one question that I've never seen raised is if there has ever been a suggestion that Centos and SL should combine, or at least work together? They seem to have exactly the same aim. I wonder why SL was set up, rather than offering to help the CentOS team? We have discussed a merger, however; they add things to the install discs that are not upstream that their users need ... we don't do that (as one example). We have different goals ... and for what SL rebuilds they want to be 100% binary compatible ... but they do not want their ISOs to necessary be compatible (if, for example, they need openais and it is not upstream). There is nothing WRONG with either approach ... they are just different. I saw statistics - I don't remember where - saying that CentOS had 30% of the Linux market, which I found very surprising, though also satsifying (to me). SL had a tiny share. (I remember now, it was someone complaining that Fedora's share was slipping badly.) http://w3techs.com/technologies/details/os-linux/all/all I belong to what may be the silent majority who don't really care if CentOS is absolutely up-to-date. (As far as I can see, none of the changes in CentOS-6.1 would make the slightest difference to me. I run CentOS on 3 home servers, and Fedora on my laptops.) I was very struck by the ease with which I upgraded to CentOS-6, compared with the nightmare (now hopefully over) upgrading from Fedora-15 to Fedora-16. It reminded me why I would never run Fedora on a server. To me, the reliability and solidity of CentOS are what I relish, and I'm very grateful to the CentOS team for their work. I don't mind them getting a bit crotchety at times! signature.asc Description: OpenPGP digital signature ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On 11/14/2011 08:56 PM, Alan McKay wrote: Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? We have a CR repository that has a bunch of 6.1 (and updates newer than 6.1 as well) in there. It is not like there are no updates to 6.0 released. The ISOs for 6.1 are not released, but the RPMs are. signature.asc Description: OpenPGP digital signature ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Johnny Hughes wrote -- We have a CR repository that has a bunch of 6.1 (and updates newer than 6.1 as well) in there. It is not like there are no updates to 6.0 released. The ISOs for 6.1 are not released, but the RPMs are. --- I was wondering if it would be safe to just stay with the 'standard' repo for centos and wait for 6.1 that way or do you suggest adding the CR repo as a necessary event? thanks. ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Tue, Nov 15, 2011 at 4:16 PM, Johnny Hughes joh...@centos.org wrote: We have different goals ... and for what SL rebuilds they want to be 100% binary compatible ... but they do not want their ISOs to necessary be compatible (if, for example, they need openais and it is not upstream). But they also include revisor, so building different ISO spins should be the least of the problems. Splitting their additions to an extras-like repo might be slightly more complicated. -- Les Mikesell lesmikes...@gmail.com ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
I was wondering if it would be safe to just stay with the 'standard' repo for centos and wait for 6.1 that way or do you suggest adding the CR repo as a necessary event? Depends on if you feel that security updates are important to your infrastructure. John -- Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed. -- Herman Melville (1819-1891), novelist and poet pgpI4R9UrBRxn.pgp Description: PGP signature ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Alan McKay wrote on 11/14/2011 09:56 PM: Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? At least partly matter of priorities. SL finally released 5.7 on 09/14/2011 and just released the LiveCD/DVDs 11/02/2011. They did provide rolling 5.x updates, analogous to CR, in the interim. CentOS went for 5.7 before 6.1. Phil ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
These seems to me to be the first message in the series and provides a really good summary of the changes at Red Hat which seem to be making life a lot more difficult for CentOS. Just figured I'd pull it out of that thread and change the subject line. Below Johnny's email I've copied another from the original thread, written by Lamar Owen, which gives some good explanation on how Red Hat is able to get away with this. Basically from what I gather, while Red Hat cannot restrict access to sources, they can restrict access to binaries. And since CentOS has a goal of binary compatibility with upstream, they are essentially left trying to hit an unknown target. But (now I'm stretching my limited knowledge even further) Scientific does not have this restriction since they are less concerned about exact binary compat. On Fri, Oct 21, 2011 at 12:54 PM, Johnny Hughes joh...@centos.org wrote: On 10/21/2011 10:01 AM, Les Mikesell wrote: On Fri, Oct 21, 2011 at 9:51 AM, Nicolas Thierry-Mieg nicolas.thierry-m...@imag.fr wrote: Johnny, chill. I don't blame him for being confused. Up until right now, you updated to a point release, then, over the weeks and months, there were updates. All of a sudden, there are *no* updates for the 6.0 point release, which is a major change in what everyone expected, based on history. this is the way it has always been: once upstream releases x.y+1 , there are no more updates to x.y (in upstream and therefore also in centos), until centos releases x.y+1 . Yes, but that used to be transparent, because the centos x.y+1 release happened quickly so it didn't matter that the update repo was held back until an iso build was done. Yes, and NOW the release process is MUCH harder. Red Hat used to have an AS release that contained everything ... we build that and we get everything. Nice and simple. Build all the packages, look at it against the AS iso set ... done. Two weeks was about as long as it took. Now, for version 6, they have: Red Hat Enterprise Linux Server (v. 6) Red Hat Enterprise Linux Workstation (v. 6) Red Hat Enterprise Linux Desktop (v. 6) Red Hat Enterprise Linux HPC Node (v. 6) Red Hat Enterprise Linux Workstation FasTrack (v. 6) Red Hat Enterprise Linux Server FasTrack (v. 6) Red Hat Enterprise Linux Desktop FasTrack (v. 6) Red Hat Enterprise Linux Scalable File System (v. 6) Red Hat Enterprise Linux Resilient Storage (v. 6) Red Hat Enterprise Linux Load Balancer (v. 6) Red Hat Enterprise Linux HPC Node FasTrack (v. 6) Red Hat Enterprise Linux High Performance Network (v. 6) Red Hat Enterprise Virtualization They have the same install groups with different packages based on the above groupings, so we have to do some kind of custom generation of the comps files to things work. They have created an optional channel in several of those groupings that is only accessible via RHN and they do not put those RPMS on any ISOs ... and they have completely changed their Authorized Use Policy so that we can NOT login to RHN and use anything that is not on a public FTP server or on an ISO set ... effectively cutting us off from the ability to check anything on the optional channel. Now we have to engineer a compilation of all those groupings, we have to figure out what parts of the optional channels go at the point release and which ones do not (the ones that are upgrades). Sometimes the only way to tell is when something does not build correctly and you have reverse an optional package to a previous version for the build, etc. We have to use anaconda to build our ISOs and upstream is using something else to build theirs .. so anaconda NEVER works anymore out of the box. We get ISOs (or usb images) that do not work and have to basically redesign anaconda. We can't look at upstream build logs, we can't get all the binary RPMs for testing and be within the Terms of Service. And with the new release, it seems that they have purposely broken the rpmmacros, and do not care to fix it: https://bugzilla.redhat.com/show_bug.cgi?id=743229 So, trust me, it is MUCH more complicated now than it was with previous releases to build. With the 5.7 release, there were several SRPMS that did not make it to the public FTP server without much prompting from us. And with the Authorized Use Policy, I can not just go to RHN and grab that SRPM and use it. If it is not public, we can no longer release it. So, the short answer is, it now takes longer. Thanks, Johnny Hughes Lamar Owen lo...@pari.edu via centos.org Oct 28 to CentOS On Friday, October 21, 2011 02:22:26 PM Les Mikesell wrote: Which is explicitly imposing additional restrictions. Which is explicitly prohibited in section 6. I don't see any exceptions relating to what the consequences of those restrictions might be. The RHN AUP simply says that if you redistribute information from RHN you lose access to RHN. It does not restrict your right to
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
On Mon, Nov 14, 2011 at 5:57 PM, Alan McKay alan.mc...@gmail.com wrote: Basically from what I gather, while Red Hat cannot restrict access to sources, they can restrict access to binaries. And since CentOS has a goal of binary compatibility with upstream, they are essentially left trying to hit an unknown target. But (now I'm stretching my limited knowledge even further) Scientific does not have this restriction since they are less concerned about exact binary compat. You are stretching your knowledge to a wrong direction :) Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. http://lists.centos.org/pipermail/centos/2011-November/119250.html Akemi ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Changes at Red Hat confouding CentOS (was: What happened to 6.1)
Both CentOS and Scientific Linux *aim* at 100% binary compatibility and they are both doing their best toward that goal. However, neither is perfect. That's interesting. So how is it they've managed to come out with 6.1 (and so long ago at that)? -- “Don't eat anything you've ever seen advertised on TV” - Michael Pollan, author of In Defense of Food ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
