Re: a year later - CERN move to Centos - what are we doing?
Konstantin Olchanski writes: > The installer for both have the same idiotic "you *must* create a fake > user or no login prompt for you!", If you're not making a local user, then you've probably got a network authentication scheme. In which case, you're probably deploying more than one machine. Take a look at making a kickstart file to automate the install process. In kickstart, you don't have to make a dummy user, you can just define your network authentication setup. And much, much more: then use that same script to install on as many machines as you want. Finally sat down to figure this out, and even for my measly handful of machines, it was worth it. > and the same "you *must* use the disk partition tool designed by > dummies for dummies". likewise solved by kickstart. although even interactively you can still use fdisk from the shell, certainly not a dumbed-down tool. Perhaps too much in the other direction :) -- Alec Habig University of Minnesota Duluth Dept. of Physics and Astronomy ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: kstars
Efraim Yawitz writes: > Why is kstars no longer part of Scientific Linux? I'm still using 5.4 > which has this wonderful and small planetarium program. Why was it > removed from later versions? doesn't directly answer your question, but I use xephem: but have had to roll my own rpm for many releases now. Just compiling a new version now as I use it for my intro astronomy course (for making current starfields etc for my lectures). Over time, non-core programs come and go from TUV repository, which composes 99% of all the packages in SL, and which the SL maintainers have no control. You can find many of the things you'd like in a supplemental repository like EPEL (unfortunately, neither kstars nor xephem): if there's a critical mass of people who want the the thing. But sometimes you just gotta do the old fashioned thing and compile it yourself :( -- Alec Habig University of Minnesota Duluth Dept. of Physics and Astronomy ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL-6.3 Install xfce -
Bob Goodwin - Zuni, Virginia, USA writes: I would like to install xfce but there doesn't seem to be a yum rpm. What is the best way to deal with that? It's there, but in the epel repository. Go get the epel-release package from here: http://fedoraproject.org/wiki/EPEL (get the right one for your arch, links midway down the page), then yum --enablerepo=epel search xfce to see what's available. EPEL has a pretty good compatability record with RHEL (that's their whole goal in life), so odds are decent this will work in your situation. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Update to buggy kernel 2.6.32-279.1.1.el6
Following up on the request-key issue, although it's sort of hijacking Daniel's thread: could be related in the sense that I wonder if using the 6.3 kernel in 6.2 isn't the source of many similar problems. In the new kernel changelog I see: * Wed Feb 29 2012 Aristeu Rozanski aroza...@redhat.com [2.6.32-245.el6] - [build] update RHEL_MINOR to '3' (Aristeu Rozanski) - [fs] keyring: allow special keyrings to be cleared (Steve Dickson) [772495] - [fs] NFS: Update idmapper documentation (Steve Dickson) [772495] - [fs] NFS: Keep idmapper include files in one place (Steve Dickson) [772495] - [fs] NFS: Fall back on old idmapper if request_key() fails (Steve Dickson) [772495] which explains why there are messages in /var/log/secure but no actual failures like the older bug report mentions. Unfortunately, this bugzilla is restricted so no more info can be gleaned. Updating a test machine to the 6.3 RC, the request-key errors go away. I interpret this as the new kernel seeing the the idmapper version in 6.2 (nfs-utils-1.2.3-15) as old but being happy with the version in 6.3 (nfs-utils-1.2.3-26). Just the sort of backport bug you'd expect by taking a 6.3 kernel and releasing it as a 6.2 update. The nfs-utils changelog does list a lot of idmapper changes on the same date: * Wed Feb 29 2012 Steve Dickson ste...@redhat.com 1.2.3-16 - Enable the keyring based idmapping (bz 772496) - Correct nfs(5) man page on how TCP retries are done (bz 737990) - Fixed sigio problem in rpc.idmapd (bz 751089) - Only link in libtirpc to binaries that need it (bz 772050) - Increase the stdio file buffer size for procfs files (bz 736741) - Fixed gssd from picking the wrong creds (bz 738774) - Mounts fail with using -o bg,vers= options (bz 740472) So: looks like updating the nfs-utils should be a requirement of updating the kernel: if that doesn't break other stuff, which I'm not in a good position to test. Question for Pat: given that I can't read the bug report in question, do I file a new one? Or is this a SL problem not a TUV issue? (not sure of the provenance of the 6.3 kernel appearing in the 6.2 updates directory). Alec Alec T. Habig writes: Daniel Kontsek writes: all of my SL 6.2 machines proposed today a kernel update to 2.6.32-279.1.1.el6, which would probably make them unreachable after reboot. This kernel is also included in the 6.3 release. Another problem which came with this kernel update is thankfully annoying rather than fatal. /var/log/secure is filling up with things like: request-key: Cannot find command to construct key 100834592 (where the int after key changes). Looks similar to: https://bugzilla.redhat.com/show_bug.cgi?format=multipleid=794780 which came into the fedora kernel when it was the version we're using in SL6 now, but which has been fixed in Fedora for a while. Anyway: if anybody else has noticed this, let me know as I try to get a proper bugzilla cooked up, so I can include your information too. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Update to buggy kernel 2.6.32-279.1.1.el6
Daniel Kontsek writes: all of my SL 6.2 machines proposed today a kernel update to 2.6.32-279.1.1.el6, which would probably make them unreachable after reboot. This kernel is also included in the 6.3 release. Another problem which came with this kernel update is thankfully annoying rather than fatal. /var/log/secure is filling up with things like: request-key: Cannot find command to construct key 100834592 (where the int after key changes). Looks similar to: https://bugzilla.redhat.com/show_bug.cgi?format=multipleid=794780 which came into the fedora kernel when it was the version we're using in SL6 now, but which has been fixed in Fedora for a while. Anyway: if anybody else has noticed this, let me know as I try to get a proper bugzilla cooked up, so I can include your information too. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: What's diff btw yum-autoupdate and yum-cron?
Winnie Lacesso writes: Whereas yum-cron contains many more config files etc, has to be chkconfig'd on, etc. The chkconfig thing was TUV policy. Packages don't come on by default w/o extensive review. Since Fedora's default package manager was not yum-cron, it wasn't on by default. Likewise SL had their own cron scripts to do this: don't want two jobs doing the same thing on by default! yum-autoupdate config file has an exclude line in it; yum-cron's doesn't. Sure it does. yum does this at a higher level than any scripts one might write: having this information in contradictory places isn't good (and has confused be before with the SL cron jobs). For example, in my laptop's /etc/yum.conf is the line: exclude=krb5-devel,krb5-workstation,krb5-libs,NetworkManager-glib,NetworkManager (NetworkManager being evil incarnate, and there being a krb5 bug I'm working around ATM). So it looks like for simplicity functionality, yum-autoupdate on SL6 gives what's wanted - can exclude things, get the daily automatic update, much less config cruft. One man's cruft is the next guy's feature (cf the above NM comment, actually). Most of what's in there were added from a server farm maintenance angle, something which matches the needs of many SL (if not Fedora) users. Whatever tool works for you, use it. I'm just happy that it's back in yum proper, and all the crazy poll for updates all the time from a desktop gui widget diversions have left the building. It's just a shame TUV has such an old version of it in v6 of the product: all I'm suggesting is if you want to use this tool and would like more flexibility with how your server collections get updated, consider downloading a newer version of the package than is available in the default repos. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: What's diff btw yum-autoupdate and yum-cron?
Konstantin Olchanski writes: I can report that all these packages are defective in different ways. (the CERN version is the least broken). Speaking as the maintainer of yum-cron before it got subsumed back into yum proper, several if not all of the bugs you mention got fixed. Note that SL6 is based on an ancient version of fedora: grabbing the yum-cron package from a newer version of fedora could help (or at lease expose more tools for your own tweaking). Hopefully we've got any yum version dependencies built into the rpms, so you'll know when you try to install if features have evolved past the point where it just won't work with the yum on your system. Sorry I can't be more specfic, but it could be worth taking a look at the package changelog to see if your issues have been addressed. And that changelog ceased to have meaning after the package rejoined yum itself, I'm afraid, as subpackages don't get their own changelogs anymore :( Looks like yum-cron-0.9.2 was the last version before assimilation, and I strongly suspect that using a current Fedora yum as a whole will plunge your system into dependancy hell (based on python versions if nothing else). Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Ghostscript problem?
Tom Rosmond writes: I did 'yum downgrades' of ImageMagick.i386 and ImageMagick.x86_64, and this corrected the problem. Has this happened to anyone else? Who else should be notified? Happened to us too. Poking around online, one gets similar errors if some of the supporting files in the ghostscript builds aren't installed correctly, so I strongly suspect a simple build or specfile problem. However, I didn't have time to debug it and went with the downgrade workaround to save time (remember to put ghostscript in your excludes list lest autoyum upgrade your downgrade!). The proper thing to do would be to check TUV's bugzilla, and if there's not already a bug report about it, to file one. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: wine on SL 5.5
Tom Rosmond writes: However, apparently the epel-release repository isn't available for 5.5. At least a 'yum install epel-release' doesn't find it. So what if any method is available to install Wine on this SL release? Is there another repository that has it, from which I can do a Yum install? It appears to be in Dag's repository. Enable /etc/yum.repos.d/dag.repo (from yum-conf-57-1.SL) and try yum search wine to see what's there. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
backing out ghostscript rpm in SL4
Hi folks, So we've got a fairly large nest of machines running SL4. Yes, we know they're EOL'd, but we're not upgrading OS's on machines actively taking data and on a non-routable subnet: we'll do it this summer during the accelerator shutdown instead. Yesterday ghostscript was updated by yum, and the new version's broken for our application (turning root created postscript into jpgs and pdfs for online monitoring jobs). Unfortunately, SL4's ancient version of yum doesn't know about downgrade, and the SL4 update repo doesn't appear to keep previous versions of rpms around so we can manually back out. So, I was wondering if anyone had the pre-yesterday version of this rpm around? ghostscript-7.07-33.13.el4 was pushed to the repo yesterday, which breaks our application. I found a completely un-updated version from centos, and downgrading to it fixes our problem: ghostscript-7.07-33.el4 but I'm hoping for something newer which gets all the update goodness except yesterday's breakage. thanks, Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: backing out ghostscript rpm in SL4
Nico Kadel-Garcia writes: This is where you get in trouble. We want all the new stuff, except the undefined stuff that doesn't work. Heh :) More concretely - whatever it was we had up until yesterday The specific details of the breakage were omitted to increase the S/N on the mailing list, and if I could compare old and new builds I'd be able to make a much more detailed bug report to TUV anyway. The ghostscript-7.07-33.11.el4 from 2009 is available under the the SL 4.8 distributations at your nearest Scientific Linux mirror. Great, thanks! Didn't think that a newer sub-version wouldn't have gotten the same fix as 4.4, but I'm not complaining and thank you for pointing this out. You can also grab the SRPMS's and take a look at which patch broke things, if you have the time. I was planning on doing this. Been spoiled by Fedora development under Koji, where grabbing most any build (even ones which never made it to the mirrors) is really useful for nailing down regressions like this. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL6x: sbin in path??
Ken Teh writes: Is there some reason why regular users get /usr/local/sbin, /sbin, /usr/sbin in their PATH? SL5x didnt do it and I believe that was also the case for SL4x. Can I remove it safely? I'm running 6.X and my regular users don't have it in their paths, with no harm done. Didn't realize the default had changed: I guess my /etc/skel stuff for making new users is just old. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Is removing selinux a bad idea?
Nico Kadel-Garcia writes: Until it breaks something, unpredictably. For example, restoration of previously working software with rsync from another working system, or tar from backup tape, will not set SELinux. The solution here is to tell selinux to please rebuild the file permissions, as the next step after restore from tape, before trying to do anything with the system: restorecon -R -vv /restored/filesystem (the verbose option is of course not necessary, just entertaining) Is selinux worth it? There are a few extra steps, but I've been living with it enabled on my systems for years and it's not too hateful. It does use extra resources on I/O to verify things are ok, but most systems aren't running so close to the perfomance edge that anyone cares. Does it help defend a system? That's a lot harder to quantify - which is a true statement for any security measure. Why? Because they're supposed to work in layers. If someone gets around one defense, the next is supposed to be there to stop them. If it's not, you're screwed. If they never make it to that next layer, you'll never know if that next layer was ever worth it. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Flash plugin
I did encounter a problem with the official adobe repo yesterday - it wanted to install the i386 version over the x86_64 version, so bombed with a file conflict. Deleting the adobe yum config rpms and relying on Dag made things work here. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Should yum-autoupdate depend on mailx?
Dennis Schridde writes: So it is indeed a bug? How should I progress from here? (I did not see a bugtracker on the website.) Since very few packages actually belong to SL (as opposed to TUV, who have their own bugzilla), we just say on the mailing list: Hey Connie and Pat, here's an easily fixed bug! Specifically, please add a mailx dependency to yum-autoupdate. Low tech but it works :) Reminds me of a Letterman cartoon episode, for example: http://www.youtube.com/watch?v=z3y_H3SaoAYfeature=results_videoplaynext=1list=PLBE024B04B63BF3E8 -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/ pgpiFpGqGEhDy.pgp Description: PGP signature
Re: SL6: how to default to KDE?
Jeff Siddall writes: So why does GDM still default to Gnome? Grrr... So I am fed up with GDM. Any ideas how I ditch GDM and switch to KDM? Here's how I did it, 3 steps: yum install -y xorg-x11-xinit-session Create a file /etc/sysconfig/desktop containing the line DISPLAYMANAGER=KDE and a patch for /etc/X11/xinit/Xsession is attached. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/ --- /etc/X11/xinit/Xsession.issue 2011-03-14 12:54:05.862817718 -0500 +++ /etc/X11/xinit/Xsession 2011-03-14 12:54:05.879720291 -0500 @@ -30,6 +30,14 @@ # Xsession and xinitrc scripts which has been factored out to avoid duplication . /etc/X11/xinit/xinitrc-common +# RHEL6 GDM doesn't seem to support selectable sessions, and always requests a +# gnome-session. So we unset this default here, to allow things like user +# .xsession or .Xclients files to be checked, and /etc/sysconfig/desktop +# settings (via /etc/X11/xinit/Xclients) honoured. +if [ -n $GDMSESSION -a $# -eq 1 -a $1 = gnome-session ]; then + shift +fi + # This Xsession.d implementation, is intended to obsolte and replace the # various mechanisms present in the 'case' statement which follows, and to # eventually be able to easily remove all hard coded window manager specific
Re: Xterm fonts; copy - lame user's questions...
Andrew Z writes: as for the copy-select - i'm on toshiba laptop and don't use the mouse. Only this annoying touchpad (trackpad? ) Do you know where i can change the strl-shift-c to ctrl-ins for copy-paste? If you're using Konsole then you're probably using KDE. Run the System Settings app, choose Keyboard and Mouse, then Standard Keyboard Shortcuts. Copy is normally ctrl-C, but has ctrl-insert as an alternate already, and Paste is ctrl-V with shift-insert as an alternate. If yours is different, you can change them right in this window. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: How to update the Kernel, keeping the proprietary modules.
Take a look at the dkms program (in the rpmforge repo). It manages a build environment which makes it easy to keep drivers rebuilt in synch with the kernel. Not sure about adaptec, but my LSI card's drivers came with all the right scripts to plug into dkms. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL vs. CentOS vs. RH EL qualification testing
Yasha Karant writes: Although the future is unclear for Fermilab with the imminent decommissioning of the Fermilab accelerator, this professional status currently is correct. Correction - one beamline (the tevatron) and associated experiments are ending. The rest of the accelerator complex and associated experiments (not to mention the non-accelerator based stuff) are humming right along, new experiments coming online, etc. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL6: NIS, AUTOFS incompatible with NetworkManager
Andy Mastbaum writes: It's fine for 3G on laptops or whatever, but that's not really the target market of SL. The average SL user's needs are different from TUV's; this seems like a place where some changes to the distribution would make some sense. Certainly one of the first things to go into a command line in a new SL install for me is rpm -e NetworkManager Then the next is to edit ifcfg-eth0 appropriately. If you're getting your network config from dhcp anyway, it's all of 4-5 lines and looks the same for most machines anyway. If you're a server admin, you've no need for a gui (or a tui) to set the /etc/sysconfig/network-scripts files options anyway - although I'll admit that I'm running on experience here and that a documented template file would be a very nice thing to have there if you don't already know what the options are. Just something with all the different variables listed but commented out, so someone with vi could go in there and uncomment or fill in what's needed for their setup. Is it worth one making one of the little SL configuration rpms which blows up NetworkManager and creates a default DHCP config, call it something like server-dhcp-networkconfig, so this can happen at install? Would be easy to add a template file to that rpm too. All of FNAL's compute cluster nodes are going to need this anyway for when they eventually go to 6.x or rolling them out will be a nightmare. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL6: NIS, AUTOFS incompatible with NetworkManager
Lamar Owen writes: I'm assuming you filed a bugzilla upstream? Also, in your ifcfg-ethX files you have the ability (that is fully documented in upstream's documentation) to tell NM to not manage a particular interface, but just pull it up. It's just a matter of learning a newer way of doing things. The big problem with this is that even if you do this, NM returns ok to the startup sequence well before it actually gets around to bringing up your network(*) - then things like autofs and nis fall over. I actually did file a bug back when this behavior hit Fedora 11 (remember, Fedora is TUV's beta cycle) and got no love (although a cleaner workaround than mine was presented): https://bugzilla.redhat.com/show_bug.cgi?id=506533 But, since this bug seems to have survived the development cycle I'm happy to reopen this particular against 6. However, I've no desire to reintroduce NM on any of my machines to verify it so would want to turn the bug over to someone who does want to fight the fight, and it also should be checked that it's still a problem in 6.1. Alec (*) - obviously not for everyone if yours works, but relying on a hardware-dependant race condition isn't a very robust solution. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL6: NIS, AUTOFS incompatible with NetworkManager
Lamar Owen writes: On Wednesday, June 08, 2011 03:24:43 PM Konstantin Olchanski wrote: This is fine if I am the only manager of the computer. [snip] It is best if one could use the tools recommended by, supplied by and document by the vendor - I want to run RHEL/SL Linux, not Konstantin's Linux. [snip] Wow. You get it. Majorly customized systems that are poorly documented are ticking timebombs; and you recognize that fact. Bravo. I agree with your assessment 100%. True only to the extent that someone with a RHEL6 cert and no other linux experience would find it easier to drop in and work on the system, which is certainly important. But the ifcfg scripts hardly qualify as a major customization - that's a fairly standard way of doing things across many distros, and in fact the only way you could do it at all until the current release (the old config gui was simply a clunky interface to these very text files). And it's all changed again on F15 (so presumably also SL7 eventually). Thanks so much for all that legwork; this is good information, and thank you for taking the time to pass it on to the list! I certainly agree with this though! -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Disk Space Utilization
James Holland writes: Don't know why this is... But check how big your other partitions are using gparted. Could it be that he's comparing the 1TB drives he's bought (which are marketed as decimal 1x10^12 bytes) with the expected (binary) 2^40 bytes? That's a 10% reduction in perceived space. If the disk format has also reserved the traditional (and now obsolete) 10% for root use only, then suddenly we're 2.5 TB down from what one would naively expect after clicking on Newegg, please send me 12 terabyte drives. gparted will show the whole capacity (ignoring this root reserve), but df won't. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: USB over IP programming experience
David McLean writes: I'm looking at trying to control various RS-232 devices over a network. I've had good luck with an old-school ethernet-serial interface from Intelligent Instrumentation: http://www.edasce.com/rs232-ethernet-interface.html It's not USB, but does do the job you describe. I'm able to plop a hub out there on the detector and control multiple RS232 devices through an ethernet connection. They provide C networking libraries which compile fine under linux. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: How to convert a standar apps, to a Service.
Pablo Cavero writes: The Only that I need is a set of bash script, like script used by the tomcat, for example, or need a special binary executable?? http://www.laliluna.de/articles/tomcat-startup-script-linux.html I want to may will use chkconfig, like an standar service. The chkconfig script part is easy, just look in /etc/rc.d/init.d, steal an existing service's script, and modify it to run what you want. Note the commented line at the top starting with chkconfig - that information specifies the run levels chkconfig will set it up to be activated in, and the startup/shutdown priority. As for the code you run, that gets trickier. All the startup scripts do is start and stop things. Those things could be more scripts, executables, or combinations thereof. Areas to pay attention to with your code (to get it to behave nicely when run this way) include: getting it daemonized properly (dropping and/or redirecting IO streams), privilege seperation (can it run as a non privleged user? if so, do that, if not, be really careful and have only the bare minimum code run with elevated privileges), logging (have it log status information, ideally with customizeable log levels, to the appropriate place), and signal handling so it shuts down cleanly when told to, reloads configuration files if asked, etc. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: xrdb gone bad. xorg-x11-server-utils-7.1-5.el5_6.1 broken?
David M. Cooke writes: Several users started complaining today about various X apps, such as xterm and emacs, that no longer look the way they want. It looks like the resources they set in their .Xresources files are no longer set. Same in EL6. The changelog for this package says: * Wed Mar 16 2011 Adam Jackson a...@redhat.com 7.4-15.el6_0.1 - cve-2011-0465: Sanitize cpp macro expansion. (CVE 2011-0465) which sounds like something that could indeed break .Xresources parsing. Although in my case, not only old-style X apps lost their fonts marbles, but so did the KDE programs, menus, etc -- which I didn't think used the old-style X fonts at all. After wasting 15 minutes resetting fonts in many different places, X is usable again. I'm sure Murphy's Law says that this bug will be fixed tomorrow and we'll all have to re-reset things :) -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL 6.0 on Thinkpad T410s crash on login.
Take a look here: http://www.thinkwiki.org/wiki/ThinkWiki linux knowledge from around the net about Thinkpads accumulates here, a very nice site. And you can log in and add your own information to be archived for future people in your situation. The T410s has Intel graphics: http://www.thinkwiki.org/wiki/Category:T410s and there's documentation of Fedora and Suse installs, with few graphics problems: http://www.thinkwiki.org/wiki/Installation_instructions_for_the_ThinkPad_T410s My general advice though - usually you're far better off going with a Fedora or Ubuntu style distro on laptops, the enterprise-based distributions update too slowly on the hardware driver front for portable computing needs. An immediate suggestion for your problem - go in in rescue mode, edit /etc/inittab, and change the default run level from 5 to 3. That way you can get in on console and tweak stuff, issuing a startx command to bring up the desktop at will instead of being forced into a buggy state right away. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: has anyone tried running an NFS network *solely* using NFSv4?
I was poking at this yesterday myself with no success, so would love to know what the answer is. This is especially important since by default, iptables is installed and active, and AFAIK the only way for nfs to coexist with iptables is use nfs4. So out of the box, nfs doesn't work unless one disables a security tool, aside from the issue that nfs4 is designed to have a much higher level of security than the older versions, such that we really should all be using it exclusively anyway. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL 6 Remote desktop login
Timmy Siu writes: I need to log in as root just through LAN. I think the general solution to this is: don't. Log in as a user, pop open an xterm, then su or sudo to do what you need to do. An extra step, but one which closes a wide variety of potential security problems: ie, a ton of GUI code which could be a security problem if you execute it as a privileged user. I'm not sure if this has propagated into SL6 or not, but current versions of Fedora default to not allowing you to log in as root even on console (in X windows mode, at least), for the same reason. And a NX session is pretty much just a virtual X session anyway, with the additional security scare of being a remotely executed giant pile of GUI code. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: problem with
By default selinux doesn't let nfs work (an intentional decision from the upstream vendor: they figure if you're savvy enough to set up nfs, you can tweak selinux to allow it). However, it's easy to change the policy to allow. Here's a blog post I googled up while trying to remember the commands involved: http://prefetch.net/blog/index.php/2010/11/02/configuring-a-linux-nfs-server-in-a-selinux-managed-environment/ Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Memory limits for Scientific Linux kernels
Stephan Wiesand writes: Just curious: why would you want to run a 32-bit OS on a machine with that much RAM? For me, binary compatability with 32bit machines in the same computing cluster. Users keeping different programs compiled differently based on which box their job happens to run on is chaotic at best, although being careful with condor can help. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: SL forums
Hi feddss, mplayer is the right answer. atrpms and dag will play nicely together if you enable the yum-priorities plugin, just yum install yum-priorities I think this is activated upon install, but if not edit /etc/yum/pluginconf.d/priorities.conf to enable it. Then you can enable both repositories and they won't step on each other. The mplayer on the machine I'm typing at appears to have come from atrpms. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Top 500 - why no SL?
Keith Lofstrom writes: What is a surprise is the nonappearance of Scientific Linux on this list: http://www.top500.org/stats/list/34/os Looking at it, if I'm correct CentOS is the only free OS on there (with 8 installs). I'm guessing that if you've got the $$ to build a top 500 computer, you're willing to pay for a service contract on your OS to better keep it running. On the other hand, what the heck is Linux. Looks like they weren't very consistent in their naming methodology. There's also Red Hat Linux by itself as well as broken into RHEL4 and RHEL5. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Resolution and (default) packages?
FRANCHISSEUR Robert writes: as long ago I enabled dag.repo and I had a lot of trouble because yum and yum-conf were updated with dag'ones which are completely different than SL ones. The situation has improved now, been using daq without problems. I suspect that the yum-priorities plugin has a lot to do with it. This establishes a priority list such that packages is multiple repositories are installed from the one with the highest priority - in this case, the SL base and updates areas. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Where do the SL users find scientific packages?
A nice pile of astronomical programs put together by ESO: http://www.eso.org/sci/data-processing/software/scisoft/ They don't have a specific RHEL or SL repo, but their FC6 one works well for me. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Updating yum-conf reactivates the nightly updates
Troy Dawson writes: It is the same for *every* version of SL. A suggestion - take a look at the .spec file from yum-cron. There's some logic in there you could steal which ensures that the current state is preserved on upgrade. Not saying the default for a fresh install should be off (as it is in yum-cron since that was in the Fedora packaging guidelines), just that upgrades should respect any changes made by the user. If someone made the decision to disable this, that decision should be respected. Alec PS - use the current Fedora yum-cron as a baseline for anything, the EPEL version is frozen at an ancient state to avoid CentOS conflicts. -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: acroread for SL3,4?
If you're comfortable with getting updates straight from Adobe rather than the distribution, Adobe does host a yum repository of their own (to which they pushed the reader v9.1 last night, in fact): rpm -ivh http://linuxdownload.adobe.com/adobe-release/adobe-release-i386-1.0-1.noarch.rpm rpm --import /etc/pki/rpm-gpg/RPM-GPG-KEY-adobe-linux yum install AdobeReader_enu They also keep the flash player for linux in there too, in which case yum install flash_plugin gets you that one. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
Re: Selinux preventing spamc from working.
John Franks writes: It appears that the new selinux-policy will not allow spamc to be run by procmail. This effectively disables spamassassin. Is there any way to fix this? I am being inundated by spam. This is SL 51 (x86_64). This is in bugs: https://bugzilla.redhat.com/show_bug.cgi?id=485107 https://bugzilla.redhat.com/show_bug.cgi?id=486187 apparently fixed upstream. Completely unclear when the new selinux package will escape into the bugfix repositories, but there's a link to a pre-release package you could try in one of those bug listings. Alec -- Alec Habig, University of Minnesota Duluth Physics Dept. ha...@neutrino.d.umn.edu http://neutrino.d.umn.edu/~habig/
