Re: Kernel 2.6.29 for F10
On Aug 18, 2009, at 5:16 AM, Reindl Harald wrote: Hi Is there a reason that since longer time 2.6.29-Kernels for Fedora 10 landing in Updates-testing followed with 2.6.27-Builds in the meantime? On machines with kmods this is a real problem because yum detetects the kmod for 2.6.27-Build and wants to install the 2.6.27 kernel which conflicts with the installed one I use the 2.6.29 since months in the meantime, F11 has it since the release and i do not realize why F10 does not get him I tried to ask this on the fedora kernel list and the moderator never approved the message. I too would like to see F10 move up to 2.6.29+ prior to end-of-life. joe -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: Lower Process Capabilities
On Jul 29, 2009, at 8:49 AM, Serge E. Hallyn wrote: The same thing can happen at the moment with capabilities for an NFS rootfs, prelink killed file caps on fedora last time I checked. Makes them useless for general purpose app protection. https://bugzilla.redhat.com/show_bug.cgi?id=456105 joe so perhaps the same solution (falling back to classic setuid if there is no selinux policy loaded) could apply? -serge -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: prelink: is it worth it?
On Jul 9, 2009, at 9:45 AM, devzero2000 wrote: 2 - not checked if this problem is actual or not: prelink erases file-based capabilities https://bugzilla.redhat.com/show_bug.cgi?id=456105 Which remains 'NEW' a year after it was opened. It was recently reconfirmed by Tomas Mraz in F11. We have a number of apps that use file system capabilities, so we don't install prelink. This is frustrating since it impacts interactive app performance. joe -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: yum 3.2.23 more than 10x slower than 3.2.21?!
On Jun 26, 2009, at 9:11 AM, Seth Vidal wrote: On Fri, 26 Jun 2009, Roberto Ragusa wrote: Hi, yum got really slower a few days ago on F10, so I investigated and found that: (3.2.23) #yum list updates real1m28.446s user1m24.553s sys 0m0.546s (3.2.21) #yum list updates real0m7.621s user0m6.673s sys 0m0.390s this is with a few repositories, no net access, no disk access (everything is cached), trivially reproduceable. No yum plugins (just downloadonly), no PackageKit installed. About 90 seconds of 100% CPU (2.4GHz Core 2) is a disaster. just for fun: set color=never in your yum.conf and run it again. No time difference here (2:23-2:26) for color=never on yum 3.2.23. More than 2 minutes of cpu time on a 3.1GHz E8400. Downgraded to 3.2.21-2 from koji and the same 'yum list updates' takes 6.4 seconds. joe -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: yum 3.2.23 more than 10x slower than 3.2.21?!
On Jun 26, 2009, at 9:49 AM, Seth Vidal wrote: On Fri, 26 Jun 2009, Joe Nall wrote: just for fun: set color=never in your yum.conf and run it again. No time difference here (2:23-2:26) for color=never on yum 3.2.23. More than 2 minutes of cpu time on a 3.1GHz E8400. Downgraded to 3.2.21-2 from koji and the same 'yum list updates' takes 6.4 seconds. and if you can make this happen every time with 3.2.23 if you could please run: time yum -d 3 list updates | grep 'time:' and paste all the output. thanks [r...@fast ~]# yum --version 3.2.21 Installed: yum-metadata-parser-1.1.2-10.fc10.x86_64 at 2009-01-04 17:01 Built: Fedora Project at 2008-10-15 13:38 Committed: James Antill james at fedoraproject.org at 2008-10-14 22:00 Installed: yum-3.2.21-2.fc10.noarch at 2009-06-26 14:42 Built: Fedora Project at 2009-01-12 17:30 Committed: Seth Vidal skvidal at fedoraproject.org at 2009-01-07 22:00 Installed: rpm-4.6.1-1.fc10.x86_64 at 2009-06-06 05:29 Built: Fedora Project at 2009-05-18 11:26 Committed: Panu Matilainen pmati...@redhat.com at 2009-05-18 22:00 [r...@fast ~]# time yum -d 3 list updates | grep 'time:' Config time: 0.067 pkgsack time: 4.729 rpmdb time: 0.000 up:Obs Init time: 0.691 up:simple updates time: 0.183 up:obs time: 0.004 up:condense time: 0.000 updates time: 1.384 real0m6.428s user0m6.004s sys 0m0.424s Second of two identical runs with 3.2.23: [r...@fast ~]# yum --version 3.2.23 Installed: rpm-4.6.1-1.fc10.x86_64 at 2009-06-06 05:29 Built: Fedora Project at 2009-05-18 11:26 Committed: Panu Matilainen pmati...@redhat.com at 2009-05-18 22:00 Installed: yum-3.2.23-3.fc10.noarch at 2009-06-26 15:02 Built: Fedora Project at 2009-05-20 22:30 Committed: Seth Vidal skvidal at fedoraproject.org at 2009-05-20 22:00 Installed: yum-metadata-parser-1.1.2-10.fc10.x86_64 at 2009-01-04 17:01 Built: Fedora Project at 2008-10-15 13:38 Committed: James Antill james at fedoraproject.org at 2008-10-14 22:00 [r...@fast ~]# time yum -d 3 list updates | grep 'time:' Config time: 0.040 pkgsack time: 141.108 rpmdb time: 0.002 up:Obs Init time: 0.158 up:simple updates time: 0.172 up:obs time: 0.005 up:condense time: 0.000 updates time: 0.846 real2m22.245s user2m21.764s sys 0m0.426s -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: yum 3.2.23 more than 10x slower than 3.2.21?!
On Jun 26, 2009, at 10:25 AM, Roberto Ragusa wrote: Seth Vidal wrote: On Fri, 26 Jun 2009, Roberto Ragusa wrote: Seth Vidal wrote: and if you can make this happen every time with 3.2.23 if you could please run: time yum -d 3 list updates | grep 'time:' You were not talking to me, but I tried and look at this: (added -C to avoid net access) # time yum -C -d 3 list updates --disablerepo=atrpms\* | grep 'time:' Config time: 0.271 repo time: 0.001 pkgsack time: 86.684 and you're POSITIVE nothing is being downloaded here? Nothing at all? Run it again, please and capture all the output and post it to a pastebin. I have a hint and it points to the same rpms in more than one repo hypothesis. yum -d 5 lists a lot of excluding for cost: firefox-3.0.11-1.fc10.i386 from updates excluding for cost: gnome-session-2.24.3-1.fc10.i386 from updates [...] I see something related to cost handling fixes in the yum Changelog. My guess is that elements are removed from an array one at a time and everything after that is moved back one position. Or some inefficiency of that kind. You may be able to reproduce it by configuring two updates repos. The performance change occurred between 3.2.21-2 and 3.2.22-5 based on testing on koji f10 downloads. joe -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: yum 3.2.23 more than 10x slower than 3.2.21?!
On Jun 26, 2009, at 10:35 AM, Seth Vidal wrote: On Fri, 26 Jun 2009, Roberto Ragusa wrote: I have a hint and it points to the same rpms in more than one repo hypothesis. yum -d 5 lists a lot of excluding for cost: firefox-3.0.11-1.fc10.i386 from updates excluding for cost: gnome-session-2.24.3-1.fc10.i386 from updates [...] I see something related to cost handling fixes in the yum Changelog. My guess is that elements are removed from an array one at a time and everything after that is moved back one position. Or some inefficiency of that kind. You may be able to reproduce it by configuring two updates repos. Grab yum from rawhide here: http://koji.fedoraproject.org/koji/buildinfo?buildID=111297 see if the recent exclude changes help the problem. Built from the src rpm because there was a python ABI change. No change. [r...@fast rpmbuild]# yum --version 3.2.23 Installed: rpm-4.6.1-1.fc10.x86_64 at 2009-06-06 05:29 Built: Fedora Project at 2009-05-18 11:26 Committed: Panu Matilainen pmati...@redhat.com at 2009-05-18 22:00 Installed: yum-3.2.23-8.fc10.noarch at 2009-06-26 15:39 Built: None at 2009-06-26 15:39 Committed: James Antill james at fedoraproject.org at 2009-06-23 00:00 Installed: yum-metadata-parser-1.1.2-10.fc10.x86_64 at 2009-01-04 17:01 Built: Fedora Project at 2008-10-15 13:38 Committed: James Antill james at fedoraproject.org at 2008-10-14 22:00 [r...@fast rpmbuild]# time yum -d 3 list updates | grep 'time:' Config time: 0.041 pkgsack time: 143.857 rpmdb time: 0.002 up:Obs Init time: 0.191 up:simple updates time: 0.410 up:obs time: 0.005 up:condense time: 0.000 updates time: 1.142 real2m25.310s user2m24.969s sys 0m0.287s -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: [Phoronix] Ubuntu 9.04 vs. Fedora 11 Performance
On Jun 12, 2009, at 11:11 AM, Rahul Sundaram wrote: On 06/12/2009 09:35 PM, Xose Vazquez Perez wrote: put selinux=0 audit=0 in kernel line at /boot/grub/grub.conf then reboot $ dmesg | egrep -i audit|selinux Kernel command line: ro root=UUID=c99c0f86-6ebc-4e0f-91ee-4a6ae7ae6aa9 vga=791 selinux=0 audit=0 audit: disabled (until reboot) SELinux: Disabled at boot. See what Torvalds says about audit and fedora kernel: http://marc.info/?l=linux-kernelm=124405016926339w=2 Turning off audit doesn't turn off SELinux. Linus is wrong about that. I think he was referring to building the kernel w/o audit. joe -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: Proposal (and yes, I'm willing to do stuff!): Must Use More Macros
On Jun 5, 2009, at 1:57 PM, Adam Williamson wrote: On Fri, 2009-06-05 at 14:40 -0400, Matthias Clasen wrote: On Fri, 2009-06-05 at 10:31 -0700, Adam Williamson wrote: It seems to me it'd make sense to convert all these kinds of snippets into macros. Am I right, or is there a reason against doing this? When this was discussed for the example of GConf schemas in the packaging committee a few weeks ago, there was quite a bit of pushback about 'obscure macros' hiding whats really going on... Honestly, that just sounds silly. It's not obscuring things, it's a sensible level of abstraction and reuse. I suspect you'd have trouble selling that position to developers - instead of calling functions from obscure external libraries, just copy and paste the code from them into every single app you build! I don't think that'd go down a storm. ;) Libraries have well defined error handling. Macros can get pretty mysterious when they start failing. Poor analogy. joe As long as there's a clear and sensible policy for how macros should be implemented (what the files should be called and what packages they should go in), they wouldn't be 'obscure' at all. All you'd need to do to check what a given macro did would be 'grep (macroname) /etc/rpm/macros.*' or something similar. -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org http://www.happyassassin.net -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: Can pungi put packages in the iso that are not installed?
On Dec 23, 2008, at 12:41 AM, Jesse Keating wrote: On Mon, 2008-12-22 at 22:29 -0600, Joe Nall wrote: Can pungi put packages/groups in the iso that are not installed? Context: I'm building spin using pungi that includes packages that get installed in a second pass in a manner similar to firstboot. The packages require a number of services to be running to properly configure themselves and can't be grouped in the initial install. I want to include the packages in the repo, but not install them until the system has more services running. Sure! Just make sure they are in the %packages section of your pungi kickstart file and pungi will make sure they're on the install tree/media. Whether or not the user can select them during the initial install is of no concern of Pungi's. Sorry, I wasn't clear. Not an interactive install. joe -- Fedora-buildsys-list mailing list Fedora-buildsys-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-buildsys-list
Re: Can pungi put packages in the iso that are not installed?
On Dec 23, 2008, at 1:55 PM, Jon Stanley wrote: On Tue, Dec 23, 2008 at 2:53 PM, Yaakov Nemoy loupgaroubl...@gmail.com wrote: Would this require a second ks file that omits the package name, for the automated install, so that it will be on the media, but not actually installed by default? The ks file that you compose from and the one that you install from aren't the same thing. Not sure what the question is. I was trying to avoid maintaining 2 ks files. I solved the problem by using sed to create an install ks from the pungi ks. joe -- Fedora-buildsys-list mailing list Fedora-buildsys-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-buildsys-list
Re: Can pungi put packages in the iso that are not installed?
On Dec 23, 2008, at 4:31 PM, Yaakov Nemoy wrote: 2008/12/23 Jesse Keating jkeat...@redhat.com: On Tue, 2008-12-23 at 14:53 -0500, Yaakov Nemoy wrote: 2008/12/23 Jesse Keating jkeat...@redhat.com: On Tue, 2008-12-23 at 08:01 -0600, Joe Nall wrote: Sorry, I wasn't clear. Not an interactive install. Doesn't matter. Put the packages in the ks file you feed pungi and they'll wind up in the compose, regardless if they get used later. Would this require a second ks file that omits the package name, for the automated install, so that it will be on the media, but not actually installed by default? There is no requirement that the ks file you use to compose is the same ks file you use to install. In fact, them being so is somewhat silly, as if you have a local mirror, just do your installs over the network, instead of doing a compose and then installing from media. I got the impression that Joe Nall was trying to use the same ks file. Hence i asked, so we could all be clear on the details. :) Old school testers like DVDs :) joe -- Fedora-buildsys-list mailing list Fedora-buildsys-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-buildsys-list
Re: Can pungi put packages in the iso that are not installed?
On Dec 23, 2008, at 3:50 PM, Jesse Keating wrote: On Tue, 2008-12-23 at 15:22 -0600, Joe Nall wrote: I was trying to avoid maintaining 2 ks files. I solved the problem by using sed to create an install ks from the pungi ks. Just curious, can I ask what output of pungi you're using and how you're doing the install? I have a package group @myapp-all that I want in the iso repo. I do not want the packages in this group prompted for or installed by anaconda. @myapp-all get installed after the first boot when some needed services are available (selinux-policy-mls and mcstransd). joe #!/bin/bash VERSION=`svnversion -n` NAME=MyApp BUILD=/var/tmp/pungi OS=$BUILD/$VERSION/`uname -i`/os/ ARGS=--destdir=$BUILD --bugurl=http://www.myapp.org/ --nosplitmedia -- nosource --name=$NAME --ver=$VERSION --force rm -rf $BUILD mkdir -p $BUILD pungi -c myapp.ks $ARGS -GCB # Customize boot menu and add remove @myapp-\* from first install cp myapp-install myapp-install.repo myapp.ks isolinux.cfg.patch $OS sed -e 's/^...@myapp-/#...@myapp-/' myapp.ks $OS/myapp.ks (cd $OS/isolinux sudo patch ../isolinux.cfg.patch) rm $OS/isolinux.cfg.patch pungi -c myapp.ks $ARGS --force -I -- Fedora-buildsys-list mailing list Fedora-buildsys-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-buildsys-list