Re: [CentOS] How to restrict reboot/poweroff from non-admins?
> Only console users (local users) are allowed to do that. It's configured > using pam (I use Centos5.8 so forgive me if this is not the same for > CentOS6). I tried to change settings in /etc/pam.d/ and that indeed works: > > /etc/pam.d/poweroff > /etc/pam.d/reboot > /etc/pam.d/halt > > I added as a second line : > auth sufficient pam_rootok.so > # prevent normal users to reboot > auth required pam_deny.so > > > But still the user locally logged on to the machine (gnome session) can > switch it off. So I think I also missed something. I can't test it right now, but reading 'man pam.d' made me wonder if 'required' in the 'auth required pam_deny.so' in the example above should be replaced with 'requisite'. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] How to restrict reboot/poweroff from non-admins?
I just noticed that CentOS (6.2) by default allows any user to reboot/poweroff system without any admin rights, or without any further questions, if using commands 'reboot' or 'poweroff'. But 'shutdown' still requires admin rights. What is the preferred way to restrict any regular user from rebooting / powering off the system (by accident)? IMHO, sudo should be required for this purpose (at least in a system with shared remote access from multiple users, single-user laptops etc may be a different case) -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Centos 5.6 3ware raid 5
> Hi, I've seen comments about the poor performance of these cards with > raid 5 configs. I have an old card with 3 x 500G IDE drives connected in > raid 5 and I'm getting around 10mb/s write performance. :-( > > I'm seeing high iowait figures at times and associated very high cpu > load average figures, probably because, under load, everything is > stacked up waiting for the disks to actually do something. > > A non-raid sata disk in the same machine manages 70mb/s > > Does anyone know what the performance is like if I used the 3ware just > as an IDE controller? I could then switch to software raid5? Since you have IDE drives, you obviously have 3ware 7506 card, which is very old. I'm not sure when it was released but that might have been before Y2k. Anyway, at least 10 years ago. It's not fair to talk about "3ware card performance" without specifying the card when you mean that old card. Their 9650/9750/9690 of today are more than one generation newer, there has been 8xxx and 9000/9500 series that also are obsolete now. If yoy use 3 drives and configure 7506 as a transparent controller for each of them, and set up a software raid, then all the drives are behind a single pci slot, which also could be a bottleneck if you don't have a 64-bit pci slot. That said, I seriously recommend jumping to something more up-to-date technology. But don't expect that newer 3ware cards were even close to that slow. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] vsftpd log issues
I have an up-to-date CentOS 6 with reasonable amount of ftp activity (a dozen of network cameras uploading images every second 24x7). The first issue was that the whole /var filesystem was about to get full, because of huge ftp daemon log. vsftpd.conf says: # You may override where the log file goes if you like. The default is shown # below. xferlog_file=/var/log/vsftpd.log Ok, the above works now. But while the setting was (by default) commented out, the default wasn't /var/log/vsftpd.log but /var/log/xferlog which was growing without limits (it was over 6 GB when I first time noticed the problem) since logrotate tried to rotate vsftpd.log -rw--- 1 root root 0 Dec 31 03:07 vsftpd.log -rw--- 1 root root 39134459 Dec 31 12:19 vsftpd.log.1 -rw--- 1 root root 433305200 Dec 30 22:03 xferlog Now, after uncommenting the log file setting line in the conf the next issue is, that logrotate does rotate the log files (the old one gets .1 postfix added to its name and a new file is created), but it still keeps writing to the original file (which is renamed now) In the ls -l listing above: - vsftpd started to write log vsftpd.log around 10pm last night (when I uncommented the log setting from the conf and restarted the daemon, until that it was logging to xferlog) - during the night logrotate has changed the name of the existing log file to ...log.1 but now, several hours later, this renamed old file is still used for logging, and the new ...log file remains empty! Is there some simple option in logrotate's conf that could change this behaviour? Or how to fix this. There must be many others who already have run into this issue. Regards, Timo ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] 6.1 Update request
>> Just ordered a Lenovo TS130. I think there are some issues with the Intel >> graphics with 6.0 and I saw where they are resolved in 6.1. Hopefully 6.1 >> can be released soon. If not, I can install Scientific Linux temporarily. >> >> Fingers crossed!! > > Or, just grab the intel xorg driver rpm from SL, and libdrm, and one other > package, the nouveau rpm. It doesn't make sense to go all the way over to > SL > if you plan on coming straight back. > I simply installed CentOS 6.0, downloaded kernel from SL6.1 repo, and installed it. Basically this is what was needed to make Intel graphics work, I think there were 1-2 other rpms I needed to upgrade too to fix dependencies, but this was easy to notice during the kernel install. This way the system automatically will roll back to pure CentOS when newer CentOS 6.1 rpms will be available. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] CentOS 6 Installation on Thinkpad X220
> Now it boots and installs. Since CentOS has an older kernel than RHEL > 6.1 I am not able to get screen resolution etc to work so I will run > RHEL on this one as Keyboard keys, Web cam, Sound, Ericsson WLAN, all > just works :) > If you have access to RHEL binaries, that's fine. I hadn't, so I installed a kernel from Scientific Linux 6.1 repo to CentOS 6.0 to make display driver work on my Acer 7550. Anyway, some day CentOS 6.1 should be available and thereafter if updates itself to genuine CentOS again... -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] resize2fs
> I am trying to resize a centos (5.2) VM drive. I use VMware and I have > increased the size of the drive by 40G. I am running resize2fs on > /dev/sdb1 (which is my root partition) but when I do I get this error: > > [root@centos ~]# resize2fs /dev/sdb1 120G > resize2fs 1.39 (29-May-2006) > The containing partition (or device) is only 19970795 (4k) blocks. > You requested a new size of 31457280 blocks. > > How can I change the block size? Yo don't need to. But to increase the filesystem size, you'll need more blocks, which means you'll first need to increase the size of the partition. Thereafter, you resize the filesystem to use that free space. So, there are two steps and now you are skipping step #1 and trying step #2 directly. Here is some help: http://www.howtoforge.com/linux_resizing_ext3_partitions -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] CentOS 6.0
>> > Can anyone pls tell me if a ThinkServer TS130 with a E3-1225 Xeon >> > processor is compatible with CentOS 6.0? I would like to >> upgrade from >> > 5.6 but my hardware will not allow me to do so. >> >> Intel >> Quad-core >> 3.10 GHz >> L2 Cache 1 MB >> L3 Cache 6 MB >> 64-bit Processing >> * Turbo Boost Technology >> * Virtualization Technology >> * Trusted Execution Technology >> * Extended Memory 64 Technology >> Platform: PC >> >> Seems OK, what specific problems are you having ? > > Sorry, I meant the hardware I have now is preventing me from upgrading to > 6.0 > > I just want to make sure CentOS 6.0 would have the necessary drivers for > the > on board video, ethernet, etc., for the TS130 before I make the purchase. Not absolutely sure about this, but I think E3-series is very close to 2nd generation i3/i5/i7. A while ago I installed CentOS 6.0 to Acer 7550 laptop that uses integrated Intel HD Graphics of i5 CPU. All I got with 6.0 kernels was vesa 1024x768 resolution. I found out 6.1 shoud fix this. Since CentOS 6.1 wasn't out then, I installed kernel from Scientific Linux 6.1 (another RHEL clone) and theafter I got very nice graphics -and high-resolution console also in runlevel 3 to my CentOS 6 laptop. So, I would say 6.0 doesn't support Intel graphics of that CPU, but 6.1 should do it. Don't know about network etc, but at least it fixes the graphics -I guess that server uses integrated graphics since they've chosen a CPU that has it. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Intel HD graphics & CentOS 6: Error inserting i915
>> >> You need at least a 2.6.38 kernel. >> >> >> >> From: http://intellinuxgraphics.org/2011Q1.html >> >> >> > >> > And in case of CentOS it means Redhat needs to backport the driver to >> > el6 >> > 2.6.32 kernel. >> >> I admit that my secret dream is that RH already would have done that for >> RHEL 6.1 kernels, otherwise it will take rather long time. AFAIK, they >> backport things like that only for new minor versions, so if it's not in >> 6.1 >> in the beginning, it won't be there before 6.2? >> > > Correct. > > You could grab the el6.1 kernel src.rpm from ftp.redhat.com and read the > changelog.. (if redhat still lists all the changes in the changelog, now > when they don't ship separate patches anymore..) Looks like it's included in RHEL 6.1: http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6/html/6.1_Release_Notes/desktop.html -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Intel HD graphics & CentOS 6: Error inserting i915
"Pasi Kärkkäinen" kirjoitti viestissä news:20110731184737.gp32...@reaktio.net... > On Sun, Jul 31, 2011 at 12:29:59PM +0200, Marc Deop wrote: >> On Saturday 30 July 2011 16:40:45 Timo Neuvonen wrote: >> > (==) Using config file: "/etc/X11/xorg.conf" >> > FATAL: Error inserting i915 >> > (/lib/modules/2.6.32-71.29.1.el6.x86_64/kernel/drivers/gpu/drm/i915/i915.ko): >> > No such device >> > (EE) No devices detected. >> >> You need at least a 2.6.38 kernel. >> >> From: http://intellinuxgraphics.org/2011Q1.html >> > > And in case of CentOS it means Redhat needs to backport the driver to el6 > 2.6.32 kernel. I admit that my secret dream is that RH already would have done that for RHEL 6.1 kernels, otherwise it will take rather long time. AFAIK, they backport things like that only for new minor versions, so if it's not in 6.1 in the beginning, it won't be there before 6.2? As a quick workaround I set up a dirty bomb: installed 2.6.38 from Fedora 15 updates repo -so far it seems to work, and I wanted to have a very CentOS-like system for testing certain things in a portable hardware (easy to carry my virtual test machine farm to home in the evening). This is just a preliminary tester, not a production system, so it won't be a big problem if it crashes some day. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Intel HD graphics & CentOS 6: Error inserting i915
I installed CentOS 6 to Acer 7750 laptop, which has Intel Sandybridge CPU ( i5-2410M) with integrated Intel HD Graphics 3000. By default X won't use intel driver for graphics, but vesa instead, which limits the resolution to 1024x768. I tried to force the use of intel driver by writing a xorg.cof file, but then X doesn't start at all. I updated libdrm and xorg-x11-drv-intel from elrepo. This intel driver 2.15 is newer than in Fedora 15 Live CD, and with it the graphics work properly. The problem seems to be related with i915 kernel module: (==) Using config file: "/etc/X11/xorg.conf" FATAL: Error inserting i915 (/lib/modules/2.6.32-71.29.1.el6.x86_64/kernel/drivers/gpu/drm/i915/i915.ko): No such device (EE) No devices detected. I've tried kernels from CentOS Base repo, and CentOS-Plus repo too, no difference. I found a similar old bugzilla entry: https://bugzilla.redhat.com/show_bug.cgi?id=495524 Somewhere (don't find the source any more) I saw that i915 would need intel_agp module, and that would be the problem. This goes a little bit beyond my knowledge level. Maybe this has been fixed somewhere about a year ago: https://patchwork.kernel.org/patch/120293/ But is this patch something that is still missing from latest CentOS kernels? If someone has a solution for this, please advice. Regards, Timo ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Upgrade
> > You didn't wait for the official release announcement ;) > > Why should he? Is he only allowed to run yum update after each > official announcement? > I still see no release announcement yet, but packages are already > available. In turn, I can see the release announcement, but no packages :-( A few more hours won't hurt me, but I'd be curious to know if I can force yum to use certain mirror that seem to be in sync. I'm located in Finland, and the two local mirros have ages of 1.8 and 1.9 days now: http://mirror-status.centos.org/ However, I can see plenty of other mirrors in nearby countries that are 3-4 hours old only. -- TiN ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] exim rpm with DKIM support for CentOS?
I haven't learned a lot about DKIM yet, but I guess I'll need to implement it in the near future, at first for signing outgoing mail. I'm using exim as an MTA. I guess the standard exim 4.63 binary rpm from CentOS repos propably does not have DKIM support yet? Or does it, maybe? Provided the standard binary rpm does not provide this functionality, what would be the most decent way to get this functionality? Does the standard source rpm already include the necessary code, re-building the binary rpm with some additional option wouldn't sound bad. If not going this way, what could be a realiable source for a suitable exim rpm (either source or both the x86 and 86_64 binaries) for use in CentOS? Regards, Timo ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] xorg-x11-drv-nouveau for CentOS?
>> Is there a way to use "Nouveau" (open source Nvidia) driver in CentOS 5 ? >> >> "nv" seems to support only the lowest resolutions thru the dvi output >> (NV44 >> chip), and I don't like the idea of installing propiertary Nvidia >> drivers. >> This way I came to Nouveau. >> >> I had a look at some Fedora rpms, but they needed something CentOS >> obviously doesn't have: >> >> # rpm -ivh xorg-x11-drv-nouveau-2.1.6-1.fc8.x86_64.rpm >> kernel-drm-nouveau = 10 is needed by >> xorg-x11-drv-nouveau-2.1.6-1.fc8.x86_64 >> > > > you can't use RPMs built for a different OS distribution, you'd need to > take the SRPM and rebuild/recompile it. Aren't the dependencies still the same? I think they are explicitly spesified in the spec file of the source rpm? ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] xorg-x11-drv-nouveau for CentOS?
Is there a way to use "Nouveau" (open source Nvidia) driver in CentOS 5 ? "nv" seems to support only the lowest resolutions thru the dvi output (NV44 chip), and I don't like the idea of installing propiertary Nvidia drivers. This way I came to Nouveau. I had a look at some Fedora rpms, but they needed something CentOS obviously doesn't have: # rpm -ivh xorg-x11-drv-nouveau-2.1.6-1.fc8.x86_64.rpm kernel-drm-nouveau = 10 is needed by xorg-x11-drv-nouveau-2.1.6-1.fc8.x86_64 Regards, Timo ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
