[opensuse] Re: Mozilla reply to list extension feature?
[EMAIL PROTECTED] wrote: [...] Because the list manager is not set up correctly to add a Reply-To header to e-mails. Don't start this silly discussion again, we had it plenty of times, just search in the archive. The list manager is not supposed to change any Reply-To header! Its current setup is correct. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel updates for new kernels
[EMAIL PROTECTED] wrote: [...] I read a lot about kernel updates in this list, but I still have the following (understanding?) problem: When I install a new kernel from on of the repositories http://download.opensuse.org/repositories/Kernel:/Vanilla/openSUSE_10.3/ or http://download.opensuse.org/repositories/Kernel:/HEAD/openSUSE_Factory/i586/ I get the most current Version of the kernel. When I now want to stick to a specific version, lets say 2.6.23 or 2.6.24 (when finished), where will I get the security updates for the version? The repositories will change to the newer branch again, won't they? Why do you want to upgrade to such a kernel? Is there a specific reason for it? You shouldn't play with those kernels unless you know exactly what you are doing or those new kernels support later hardware that isn't supported by the box kernel. Stick with the box kernel, and you will get all necessary updates. The repositories mentioned above (and most other kernel repositories) contain snapshots of kernels as RPMs. There's no archive for those kernel RPMs and there are no updates. Those kernels aren't officially supported. Also, where do I get the 2.6.23 build now, when there are already the 2.6.24-rc builds in the repositories? You don't get them anymore. The repositories contain snapshots only. If you want to compile your own kernel, you don't have to create a kernel RPM package. There's no real advantage in doing so. However, be warned: when you compile your own kernel, then it's up to you to install the latest patches/bugfixes/updates to keep your system secure. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Enigmail header problem in Thunderbird on SuSE...
Carlos E. R. wrote: [...] The only thing I know is that I'm using only opensuse rpms, and they have this problem. I haven't tried external sources of thunderbird or enigmail. To trigger it, I only have to use local chars, and this email will not verify with enigmail: áéíóúéUýUñ¿?¡! I didn't read the whole thread, but my Thunderbird (with enigmail) certainly tells me that your email has an untrusted good signature. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Tool to extract/save mail attachments?
Anders Norrbring wrote: Anyone aware of some tool with which I can go through an entire IMAP folder and save ALL e-mail attachments to files? It's a bit annoying to go to each mail, click attachments and save individually... (On a sidenote, there are about 4000 e-mails, and around 6-10 attachments to each mail...) What about this extension for Thunderbird? https://addons.mozilla.org/en-US/thunderbird/addon/556 However, I don't know whether it works with IMAP. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel upgrade question
Constant Brouerius van nidek wrote: [...] Is information about vanilla kernels also available via a smart repository? Just to see regularly what is brewing and if some time is available to try it out? There is no smart repository. RTFM: http://labix.org/smart/faq#head-3b6a7533dee5f87959c30ca819e11d155697b275 Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel upgrade question
Linda Walsh wrote: [...] If you are wanting to play with current vanilla kernels, though, you'll have to learn how download, extract, configure and build. Not necessarily. Vanilla kernels are available as RPMs through the opensuse build service: http://download.opensuse.org/repositories/Kernel:/Vanilla/ For instance, you'll be able to install the current Vanilla 2.6.24-rc6 using this repository. However, this service is available for SuSE 10.2 and 10.3 only. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Where to find Linux header files development package ?
Carlos E. R. wrote: The Sunday 2007-12-02 at 06:52 -0600, Rajko M. wrote: Carlos: make mrproper is valid target if you want to remove all traces of previous compilation(s) and start from scratch, if it breaks something it is better to know at that stage than later when kernel doesn't work properly. It is a valid target when you get the sources elsewhere, add bits and ends, etc. No. It's a valid Makefile target for all Linux kernels, including SuSE kernels. It is not needed if you just install the source rpm from suse, which comes already cleaned. It can delete customizations, too: the only time I used it deleted my own config files and notes, and some of suse's files. I had to reinstall the sources. It deletes *anything* it thinks doesn't belong on the kernel. No. It deletes everything that can be reproduced, including a .config, a version.h, and utsrelease.h. That's the purpose of this Makefile target. After running a make mrproper, the source tree is in a clean state. There's no need to reinstall the source, all Linux kernel source is still on disk (however, the kernel source is unconfigured at that point). If you need to clean a previous compilation, use make clean. That's more than enough. No. For instance, configure the kernel source in the source tree. Later on, try to use a build directory. It won't work and the kernel build system will tell you something like this: make -C /kernel/source O=/kernel/build Using /kernel/source as source for kernel /kernel/source is not clean, please run 'make mrproper' in the '/kernel/source' directory. I have been making my own kernels for years and I never use mrproper. SuSE does not recomend it in their own instructions. It's not even mentioned. So what? Many things aren't mentioned in the SuSE documentation, but nevertheless those things can be quite useful. I use make mrproper a lot. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Where to find Linux header files development package ?
Carlos E. R. wrote: The Monday 2007-12-03 at 19:43 -, Thomas Hertweck wrote: ... No. For instance, configure the kernel source in the source tree. Later on, try to use a build directory. It won't work and the kernel build system will tell you something like this: That's a special case and not a typical use. You are not doing a make alone. It's nothing special. Just dealing with various kernels - with and without build directory - in different situations on a regular basis. I have been making my own kernels for years and I never use mrproper. SuSE does not recomend it in their own instructions. It's not even mentioned. So what? Many things aren't mentioned in the SuSE documentation, but nevertheless those things can be quite useful. I use make mrproper a lot. And I never use it. No need. I was bitten by it once, no more, thanks. Well, it seems as if you had trouble understanding the purpose of make mrproper judging from your comments that you were surprised about the disappearance of .config. If you were dealing with build directories and various kernel trees etc., then you would probably see how useful this Makefile target can be, and you would also know that a reinstallation of the source had not been necessary. Please don't recommend not to use this Makefile target solely based on your own bad experience where you were bitten once - this was obviously your own fault, not knowing what make mrproper is all about at that time. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] OpenSuse 10.2 - Fortran compilation very slow through NFS network with a 64bit server and 32bit clients.
Michel Rasquin wrote: I am a new OpenSuSE user and quite convinced by this distribution. However, I still have a problem. I have configured a gigabit network, using nfs and everything works fine except for all the Fortran compilers I have tried so far (Intel, g95, gfortran). There is no problem to compile some files which are stored locally on a nfs client. However, when I try to compile some files stored on /home which is mounted through nfs, then it takes hours to make the link between all the .o files created. That's not surprising, happens here as well and has nothing to do with the programming language Fortran etc. Don't put object files on NFS disks. You could store your source code on a network disk if you want to share it between hosts, but the build directory should always be on a local disk, in particular if you want to use multi-file optimization. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Atom:Bug-Buddy?
Josef Assad wrote: [...] Just so you don't feel all alone, I've seen this one also. Will try to reproduce and troubleshoot tomorrow, almighty schedule-deity permitting. https://bugzilla.novell.com/show_bug.cgi?id=342143 Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to tell SuSE folks to NOT include yealink driver in their distro
Chee How Chua wrote: echo install yealink /bin/true /etc/modprobe.d/yealink should help with that too. Can you kindly explain what effect this line has? I know that it will create/append a file named 'yealink' under /etc/modprobe.d, and that this file will contain the line 'install yealink /bin/true'. But what will happen when you do this? You should read man modprobe.conf. An install primitive like the one mentioned above in modprobe's configuration file will tell modprobe to run the specified command /bin/true instead of actually inserting the module in the kernel. In other words, whenever modprobe receives the command to load the yealink module, it will run /bin/true instead which always returns successful. The solution presented above will therefore effectively prevent the loading of the yealink kernel module via modprobe. However, using insmod might still work for the yealink module. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How do I manage kernels in opensuse
Rajko M. wrote: [...] Whatever method you use check symlinks /boot/vmlinuz for kernel, /boot/initrd for initrd. and /usr/src/linux and /usr/src/linux-obj for sources. The last two links should not be required. In fact, these links should not be used at all. Use /lib/modules/`uname -r`/{build,source}. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kdar for 10.3
russbucket wrote: [...] Is there a program in 10.3 to replace kdar? I use kdar in 10.2 to backup my /home and other files. K3B does not do this well. Istvan Gabor wrote: [...] Is there any alternative frontend to dar similar to kdar? That possibly can even handle backups created by kdar? Sorry, I can't answer these questions. The problem with kdar was discussed on opensuse-de, but I don't recall whether any alternatives have been mentioned. However, you could perhaps follow the following advice: http://www.linuxquestions.org/questions/susenovell-60/kdar-for-opensuse-10.3-595266/ Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kdar for 10.3
Istvan Gabor wrote: Where can I find kdar package for openSUSE 10.3? As far as I know, kdar was dropped from the 10.3 distribution since it has not been maintained upstream for quite some time and the existing version of kdar does no longer work with the dar version that is shipped with 10.3. Maybe you can find some patched kdar RPMs on the net... Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 10.3 and swap question on 4-GB ram
Bryen wrote: [...] Coincidentally, I just happened to be reading up on SWAP only moments ago in a study guide. In my material, it recommends instead of doubling SWAP to the size of RAM, you should spread SWAP into partitions in multiple disks. [...] I'm only paraphrasing what I was just reading, but it makes sense to me. If anyone else has had experience with this type of optimization, would love to hear more from you! This is nothing new. It's also mentioned (although not as obvious as it could be) in the man page: Swap pages are allocated from areas in priority order, highest priority first. For areas with different priorities, a higher-priority area is exhausted before using a lower-priority area. If two or more areas have the same priority, and it is the highest priority available, pages are allocated on a round-robin basis between them. (man 2 swapon) For further swap tuning, you should also have a look at swapiness. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to get list of variables
Bryen wrote: How do I get a complete listing of available variables in bash? Chapter PARAMETERS, section Shell Variables in man bash. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to get list of variables
Randall R Schulz wrote: [...] I use this shell procedure (defined in my .bashrc) to query local variables: vq () { set | egrep -e $1 } Isn't set also reporting functions etc.? I usually use printenv to query (environment) variables... Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 10.3 and swap question on 4-GB ram
Robert Lewis wrote: I helped a friend install 10.3 64-bit on his Q6600 machine that has 4-GB of ram yesterday. Wow was the installation fast. SUSE setup a 2-GB swap space by default. We over road that and made it 4-GB. Why? If you really need that much virtual memory, you should upgrade your RAM. It doesn't make sense to have such huge swap partitions, your system will be unusable if you need 4 GB swap (well, sort of unusable). And if you really need it at some point, you can always make a swap file which is almost as fast as a swap partition. If you want to use suspend to disk, then of course your swap space should be large enough. Is this a bug/oversite or on purpose? Why should it be a bug? The times when swap partitions had to be as big (or bigger) as RAM size are long gone. If on purpose what is the logic behind that? There's no need for huge swap partitions unless you want to use suspend to disk. And that's unlikely for a machine that seems to be used as a server. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Command line list of kill sigs
Bryen wrote: How can I pull up a list of kill or killall signals from the command line? man 7 signal Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 10.3 and swap question on 4-GB ram
Robert Lewis wrote: [...] The reason both of us did this is we came from a world in Linux where one always made swap the size of RAM or larger to allow for later ram expansion. That doesn't make sense nowadays. When you add more RAM, then you usually need less swap. As I said before, the times when swap space had to be as big as (or even double) the RAM size are long gone. I agree swapping to a file is a good way to expand swap down stream. It's an easy way to provide enough virtual memory in those rare circumstances where it might be required. Again, in the good old times swap files where a lot slower than swap partitions, but this does no longer hold with kernel 2.6. How did SUSE decide to set it to 2-GB and why? Well, look at the source code of the installation program to find out why it has chosen 2 GB of swap. Maybe it's the upper limit. Having more than that usually does not make much sense. Is there any harm doing what we did? There's no harm, it's just a waste of disk space. Th. PS. I don't need and I don't want private copies of list emails! Why would somebody want to receive all emails twice? -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 10.3 and swap question on 4-GB ram
jdd wrote: [...] I still don't see what is the problem. The disk space is cheap and if the system don't need swap, it won't use it, so I see only a vaste of disk space... That much swap space doesn't make sense. If you are happy doing things that don't make sense and want to waste your disk space that's up to you. But others prefer wiser decision making. So did the installation tool. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] pthread compiling problem
Donato Azevedo wrote: I am having a problem compiling several packages because of a pthreads include problem. When I try and compile a wxWidgets app or rta (www.rtai.org) I get the following error: /usr/local/include/pthread.h:285: error: conflicting declaration [...] from what Ive seen pthread.h is bundled with glibc-devel package Correct. But the standard pthread.h (from glibc-devel) is in /usr/include, not in /usr/local/include. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Thunderbird password nuisance.
Carlos E. R. wrote: I have a nuisance in Thunderbird. When I try to send an email via my provider smtp server, it discovers it needs a password and asks for my master password. I type it, but by this time the smtp session timed out in the server (I guess) and fails. It can't be a timeout unless you need a very long time to enter your master password. If you didn't store your SMTP password, then you would have to enter it (instead of your TB master password) each time you want to send emails, and the provider must allow for some time to do so. I try again, and this time Th. asks for my smtp password and whether I want to keep it. Why? He already knows that! This is a nuisance, because I use random passwords difficult to remember and have look it up. The reason for this is most likely that your provider had a database problem and the password was rejected (I experienced it with my provider as well, but rarely). In this case, TB asks you for a new SMTP password since it came to the conclusion that the stored one was wrong - from TB's point of view, it does not make sense to ask you for the master password again or to offer a retry button under these circumstances. Is there a way to make Thunderbird remember my password for ever and not ask for it again, even if it fails? It depends on how it fails. When there's a network problem, TB does not ask you again. It's your provider's server that most likely returned an error code indicating that the password was wrong or failed. I want to be able to enter the password in the configuration, not as a pop up when I want to send or receive. I do not want to be asked again if it fails, because once configured the password it keeps being correct. As a workaround, press the Cancel button when TB asks for the SMTP password (there might some other windows popping up, just Cancel all these windows). Once the windows are gone, try to connect again (Get Mail) and at that time, TB will remember your stored password. If the problem at the provider's end was only temporarily, it should now work. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Re: RAM
Jeremy Figgins wrote: [...] I'm sitting right now in front of my 1gig machine. I have my normal set of apps open: firefox, thunderbird, konsole, etc. How can I tell if I would benefit from additional RAM? What command and what output do I need to pay attention to? You've already seen many replies to your question. Most of them focus on paging and swapping (and actual swap usage, but that's not a good idea). When there are a lot of these activities on your system, then you could obviously benefit from more RAM. However, even without these activities you could probably benefit from more RAM because the additional RAM can be used as cache. Unfortunately, this is hard to measure. In general, if data are already in cache and data don't have to be read from disk (again), then things go a lot faster. Disk access is slow and data throughput is low compared to memory access and throughput. That's why the overall performance of your system from a user's perspective can be better with a 2.6GHz CPU and more RAM instead of a 3.2GHz CPU and less RAM. But be careful, this observation does not hold in all circumstances (depends on how much RAM there is, the use of the system itself: storage server, compute server, desktop system, etc.) and one needs to check each individual situation... Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Re: RAM
JJB wrote: Price for the system is the same with quad core 1.6 or dual core 3.0 ghz, It has mirrored drives... Do your think that cvs + subversion will get any benefit from 2gb vs 1gb? I think if your new server is only going to be used as SVN server for 15 developers and nothing else (it's not a public server etc.), then it's certainly not necessary to have lots of CPU power. It won't be used, even if you stored lots of binary data or had fancy hook scripts. I would consider a decent network connection, fast disks and a good backup mechanism as more relevant than CPU power. As others mentioned, it's difficult to answer general questions like yours because there's not enough information given. In general, I would consider an increase in RAM size more worthwhile than an increase in CPU power for any system that is not going to be CPU-bound anyway or used in HPC environments. I know of people who bought a P4 3.0 GHz with 512 MB RAM and told me that they couldn't afford to buy more RAM because of their budget limitation. However, it doesn't make sense - if they bought a 2.6 GHz P4 and 2 GB RAM, they would have been better off, for the same price... Th. PS. Why do you want to use SVN and CVS? You can import CVS into SVN repositories. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Re: RAM
JJB wrote: This server will be used by a team of about 15 developers for CVS and Subversion. [...] It is a quad-core xeon Dell machine. You want to use a quad-core system with 1GB RAM as SVN server for 15 developers? That's a bit odd from my point of view. I would rather use a standard dual-core CPU, 2GB of RAM, and a RAID system with two large disks (mirrored). Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] KDE newbie - performance monitoring?
Greg Freemyer wrote: I use ps, top, vmstat, iostat, etc. all the time from the CLI. What's available for performance monitoring via GUI? xosview Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Ext3 vs Reiser FS
Peter Sjoberg wrote: On Tue, 2007-09-11 at 09:01 +0100, G T Smith wrote: As for ext3 until someone fixes the truly appalling performance of the consistency checks (and the default behaviour of firing up those checks periodically), I am sticking with reiser. The checks can be disabled (or the interval can be changed) via tune2fs. That's one thing that really annoys me. A while ago I changed a production system with about 1.5T storage and since it been over 180 days since last fsck my reboot took almost 1 hour (when I had told the client that a reboot is fast, takes max 15 minutes) Experienced admins use the fastboot option under such circumstances. Regards, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel updates
Roger Oberholtzer wrote: I know that this is a topic that gets endless questions. Perhaps because a mistake can be so problematic. Here goes mine: I am running an openSUSE 10.0 system with kernel 2.6.13-15.16-default. This is a rather recent security-fixed kernel. But still the 2.6.13 that came with 10.0. I am looking at trying a more recent kernel. Like kernel-default-2.6.22.5-4.1.i586.rpm, which is available from openSUSE's downloads. A few questions: 1) This contains all the various SUSE patches so it will (or should) fit in like the 2.6.13 kernel did? I am not sure what you mean by fit in. It will most likely require updates of other packages to work correctly. RPM will probably tell you that it requires later versions of some packages when you try to install the kernel RPM. 2) Will this keep my current kernel? Meaning, can this rpm be removed allowing the original kernel to be used? The SuSE kernel RPMs will change your bootloader config via the RPM postinstall script. In principle, you can install several kernels in parallel (don't use RPM's -U option). All you have to make sure is that your bootloader config contains correct entries for both kernels (after the installation) so you can decide which one to boot at the grub boot menu. The installation of the new kernel RPM will probably change the default kernel entry in grub's menu.conf and if you're unlucky it will also mess up some old entries (I've seen it happening several times; it will probably also change some links in /boot) - therefore, you should check the bootloader config and /boot before rebooting the system. I am usually compiling and installing new kernels myself, I do not use the RPM packages. 3) Are there any issues with glibc when going from 2.6.13 to 2.6.22? Or any other issues? I see that sysfsutils needs updating. I guess this is because of the disk name changes. However, I do not find any sysfsutils for 10.0 when searching openUSUE (http://software.opensuse.org/search). The same is true for irqbalance. You not only have to install a new kernel but other software as well (you've already mentioned it), and this software needs to be backward-compatible otherwise it could happen that you're no longer able to run the old kernel (usually, it is). The main problem is that you might have to compile all of these additional packages yourself because nobody offers ready-to-use RPMs for old distro versions like SuSE 10.0. And sometimes these software packages require updates of other software packages, and so on. Sometimes it can become quite tricky... My recommendation: think twice whether you really need a new kernel. I am also using 10.0 here and the kernel works fine for me. Do you need some of the new features in the latest kernel? If you only want to have a look at it and play with the latest kernel version, then it might be easier to setup a new openSuSE 10.2 or beta 10.3 on a spare partition (or in a virtual environment) and use it there. Otherwise you have to go through the process of updating some packages on 10.0 - whether this is easy or not depends a bit on your experience. Take care, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] eth0 become eth1 after upgraded
Benji Weber wrote: On 26/08/07, Thomas Hertweck [EMAIL PROTECTED] wrote: No, I expect it to be the same when performing an update (as we did) - if an update causes problems like that, something has gone wrong and the distributor did a bad job. An update is not allowed to change any critical system settings. It's not a critical system setting, it may change just on a reboot, if you change nothing... You are a funny guy, right? Have you ever worked in an environment where Linux systems are actually used in production (large-scale)? I guess the obvious answer is no, otherwise you would know that these settings are critical. They are not allowed to change during or after a simple update (patch), certainly not in an enterprise version which (by definition) should provide a stable environment. Sorry, but sometimes I can only shake my head when reading these overly naive statements. Geez, it's time to end this thread... PS: Please do NOT send private copies of list emails. I am reading this mailing list - why do you think I want to receive your emails twice? -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] eth0 become eth1 after upgraded
Benji Weber wrote: [...] My point is this particular setting might just change after a reboot, it might have nothing to do with a patch. There are persistent names you can rely on, or you can choose to gamble that the ethX names that are transient may remain the same after a reboot or reloading the driver. You still don't get it. But I don't have time to explain it over and over again. If you don't understand the purpose of an update and why it's important that no settings change during an update, then it's a waste of time anyway. I'll remind you of this thread next time you run YOU and something stops working afterwards... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] eth0 become eth1 after upgraded
Jan Engelhardt wrote: On Aug 26 2007 16:54, Zhang Weiwu wrote: [...] Seems the on-board card /was/ eth0, it's getting renamed to eth1 for some reason. Eh well. What I can think of: you previously had a PCI network card. Since cards on the PCI bus are usually detected before any on-board stuff (rightfully so), eth0 is your PCI card, and eth1 is onboard. SUSE then makes sure this is the case on every boot, even if you remove the PCI one. Does that apply? By the way: we had the same problem when we upgraded from RHEL4.4 to RHEL4.5. No hardware has been changed during this upgrade. See also [1] for a similar problem. It's a bit more general and not only opensuse related. Th. [1]http://groups.google.com/group/linux.debian.user/browse_thread/thread/653281e5e81d1dd8/200fd432876126c0 -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] eth0 become eth1 after upgraded
Benji Weber wrote: [...] You shouldn't really rely on the ethX numbering remaining the same, they can change at any time. No, I expect it to be the same when performing an update (as we did) - if an update causes problems like that, something has gone wrong and the distributor did a bad job. An update is not allowed to change any critical system settings. The situation is different when you perform an upgrade, i.e. you install a complete new version of a distribution... Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] how to use this with thunderbird?
Cristian Rodriguez wrote: Jacky Woo escribió: I got millions of emails these days and no idea how to reply... How to do it? Use the reply-to-all button and remove the excess recipient. this lists unfortunately have a behaviuor that no massively deployed, real life email client supports. Stop spreading this nonsense! KMail, Thunderbird (with extension), mutt, etc. all have a reply-to-list functionality. If you don't use it, then don't blame the MUA or the list admin or the list configuration... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] how to use this with thunderbird?
Cristian Rodriguez wrote: [...] hello ? I said massively deployed clients!! (aka. Outlook, Thunderbird) also, the Thunderbird extension is not included by defualt and requires a patched thunderbird version + enigmail. Hello? Are you really working for SuSE or did you win this email address in a lottery? By all means, you should know that SuSE's Thunderbird RPMs have all patches etc. included (I've used a list-reply to write and send this email). Where's your problem? This is an opensuse mailing list, so we would expect to mainly see people here using the opensuse distro, or planning to use opensuse - and yes, typical mail clients included in this distro support a list-reply functionality. If you prefer to ignore this, or if you you think that the massively deployed Windows MUA Outlook is relevant enough to change the configuration of this list, then it's your fault. End of story. I am really happy that the list admin does not follow your suggestions. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel choices
ianseeks wrote: [...] If you were going to make a choice to download the latest (or as close to latest) kernel from the repositories - which path do you take? If i go to this mirror for example http://ftp-1.gwdg.de/pub/opensuse/repositories/Kernel:/ do i go to HEAD/openSUSE_Factory/i586/ or Vanilla/SUSE_Factory/ Well, I don't know the details, but the name vanilla implies that this is the vanilla kernel from kernel.org while the HEAD directory contains the SuSE kernel and kernel sources (i.e. vanilla kernel with additional SuSE patches). Therefore, the answer to your question is: you need to decide which kernel you would like to use and then choose the corresponding directory. In any case, do not replace your existing kernel, install any new kernel in addition to the old one. HTH, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel building and make xconfig
Damon Register wrote: [...] docs 2 and 3 cover the idea of keeping the old kernel along with the new one in case the new one doesn't work but it isn't very clear for me. Yes, I recommend this as well. When you compile your own kernel, it's very easy to do that. All you have to do is follow some simple rules. [...] Is there some really easy way to keep the old? I vaguely remember something about copying /boot/vmlinuz. Do I copy that and the /boot/initrd-x.img files then tamper with /boot/grub/menu.lst to include the old one? [...] (5) Install the kernel and the modules (``make modules_install'', followed by ``make install''). I would not recommend using make install for various reasons (I made some bad experiences with this command). It's really simple to do it by hand: first of all, you have to make sure that the kernel you are going to compile has a unique kernel version. With SuSE kernel sources, you can do that at configuration time, you don't have to change the kernel Makefile. Simply add a local version string under general setup (I hope this is still the correct way to do it, I have no 10.2 here, only a SuSE 10.0 system). Then compile the kernel and the modules. All of this can be done as normal user. To install the modules and the kernel, you have to become root. As root, use make modules_install to install the kernel modules. This will create a directory under /lib/modules, e.g. 2.6.20.2-local (here, you can see the string that has been added as mentioned above to avoid a clash with any default SuSE kernel or any other kernel on the system). Then you have to copy the kernel image and System.map, here an example for i386 architecture (you're in the top-level directory of the kernel sources or in the build directory if you used one for the kernel compilation): $ cp ./arch/i386/boot/bzImage /boot/vmlinuz-2.6.20.2-local $ cp ./System.map /boot/System.map-2.6.20.2-local Finally, you have to create an initial ramdisk. The actual command to do that will vary depending on whether you want a bootsplash etc., see the mkinitrd help for details. $ cd /boot $ mkinitrd -k vmlinuz-2.6.20.2-local -i initrd-2.6.20.2-local And last but not least, you have to add the new kernel to your bootloader config. If you have an existing section in grub's menu.lst like title linux root (hd0,5) kernel /boot/vmlinuz root=/dev/hda6 vga=0x342 resume=/dev/hda7 initrd /boot/initrd then just copy that and make the necessary changes: title linux.new root (hd0,5) kernel /boot/vmlinuz-2.6.20.2-local root=/dev/hda6 vga=0x342 resume=/dev/hda7 == this should be at the end of the previous line initrd /boot/initrd-2.6.20.2-local Of course, this here is only an example which will most likely not work on your system. That's all. Now you should reboot and select the new kernel in the grub menu. If you configured your kernel correctly and if there are no other problems (you've installed the kernel correctly and you made all necessary changes to menu.lst, etc.) then the new kernel should boot successfully. If you used any third-party closed-source kernel modules with the old kernel, then you have to recompile those modules for the new kernel. If one of those modules is required to boot the system, then you have to do that before rebooting the system and add it to initrd; otherwise booting the new kernel will end with a kernel panic. However, if you followed the steps mentioned above, you would then be able to boot the old kernel again to solve the problems... ;-) HTH, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel building and make xconfig
Damon Register wrote: [...] Interested in translating some day? Well, writing this howto in German was intentional. Usually, you find quite a lot of information about the kernel in English but only few readable[*] documentation in German. Many people on the German suse-linux list complained about this fact, so I decided in 2002 to write an up-to-date kernel howto in German. In principle, I could translate it but, well, it's a lot of work and I am not sure whether it's worth doing. As mentioned above, there is quite a lot of stuff available in English - it's maybe not SuSE-specific. However, for the SuSE-specific part you can always refer to Andreas Gruenbacher's documentation which is included in the SuSE kernel sources. [...] The only thing I haven't figured out at this point is why the previous discussion talked about make modules and make modules_install but didn't seem to cover building the kernel. Did I miss something there? Was Randall intending to build only the modules? Yes, the discussion was about re-enabling USB_DEVICEFS which had been disabled in the SuSE 10.2 standard kernel at that time. USB_DEVICEFS can be enabled by recompiling the usbcore module, so we focused on modules, not on building a complete kernel. Regards, Th. [*]readable in this context means: the documentation is not simplistic but also not too complex and technical; it should focus on the practical aspects a user is interested in... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel building and make xconfig
Damon Register wrote: I was just reading in the SuSE Admin Guide for 9.1 about the kernel compiling procedure. I read something that caught my attention because I didn't notice it before and I don't understand it. In section 10.3.3 it says As the configuration with make xconfig is not as well maintained as the other configuration possibilites, run the command make oldconfig after using this configuration method. Can anyone explain this? What is missing from xconfig and what does the make oldconfig do to correct this? I think SuSE 9.1 was the first SuSE distribution which came with kernel 2.6. Due to the change of the kernel build system (kernel 2.4 used a different way to configure and compile the kernel), the make xconfig part of the kernel build system was not very well maintained in early 2.6 kernels. As far as I know, developers had focussed on the make config or make menuconfig part. It could be that above mentioned comment in the Admin Guide is related to these circumstances. However, I do not recall any details - SuSE 9.1, that's a long time ago ;-) I tried to find a more recent manual or howto that covers kernel building but I don't find any at http://www.opensuse.org. Can anyone please tell me where I can find some current instruction on kernel building? http://www.digitalhermit.com/linux/Kernel-Build-HOWTO.html This Howto replaced the old The Linux Kernel Howto. It's not SuSE specific. I wrote a SuSE-specific Kernel Howto, but it's only available in German at the moment, see http://lists.opensuse.org/opensuse/2007-03/msg01324.html for details. There's also an official SuSE guide written by Andreas Gruenbacher, see http://www.suse.de/~agruen/kernel-doc/ for details (this document is also part of the SuSE kernel sources). I also noticed in my 9.1 book, the method to use before configuring the kernel is zcat /proc/config.gz .config make oldconfig I noticed that make cloneconfig from earlier instructions still works but it is not mentioned in the 9.1 instructions. Are the results different? No, the results should be the same. However, make cloneconfig is only available with SuSE kernel sources. The Makefile of the Vanilla kernel sources does not know about this target. HTH, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] bugzilla.novell.com ?
Per Jessen wrote: Has anyone succeeded in accessing bugzilla this weekend? I've been trying all afternoon, and it's not really responding. Most likely related to this message: June 16, 2007 - Bugzilla 3.0 We have updated to the latest version of Bugzilla. Here are a few of the new features. [...] Should work by now. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] New Subject - building clusters with OpenSuse
Michael Folsom wrote: I checked Science and Productivity and its not there - However, before we do that - it might be worthwhile to consider what else do we need for clustering? It would be really great to see OpenSuse shipped on clusters of all sizes from the little guys to biggies but to make that easy the missing parts should be supplied. I think building, running, and maintaining an HPC Linux cluster requires more than just a collection of software tools. Anyway... Here's my first cut at a list... - hearbeat (not sure its totally necessary but should be semi-easy) - OpenPBS http://www.openpbs.org/ - Oscar http://oscar.openclustergroup.org/ - OpenMPI http://www.open-mpi.org/ http://ganglia.sourceforge.net/ (might be part of oscar) http://munin.projects.linpro.no/ http://www.clusterresources.com/pages/products/torque-resource-manager.php http://www.csm.ornl.gov/torc/C3/ ... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] SUSE 10.0: How to prevent mount of USB disk with subfs?
Joachim Schrod wrote: Somehow SUSE 10.0 is set up to mount USB disk devices automatically as subfs on /media/* where * is the disk name. While that has been OK so far for USB sticks (I don't use them often anyhow), it is not what I want for my shiny new 500GB USB disk that I want to use for videos. I would like to prevent that disk to be mounted via the subfs mechanism, but want to use my own mount parameters. Actually, I don't want to turn off the automatism completely, that's convenient for USB sticks. I just want to turn it off for this one disk. Therefore, how can I do that? I have no problem in mounting and/or umounting this disk manually (it will be on all the time, so I can simply add it to fstab to handle it during boot/shutdown). $ cat /usr/share/hal/fdi/policy/20thirdparty/no_usbdisk_automount.fdi ?xml version=1.0 encoding=ISO-8859-1? !-- -*- SGML -*- -- deviceinfo version=0.2 device match key=storage.drive_type string=disk match key=storage.bus string=usb match key=info.vendor string=WDC WD30 match key=info.product string=00JB-00KFA0 merge key=storage.policy.should_mount type=boolfalse/merge /match /match /match /match /device /deviceinfo Vendor and product strings need to be changed to match your 500GB USB disk (see lshal for details). Do not forget to restart HAL. Hope this helps, Thomas -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] opensuse repositories now illegal in Germany
Michael S. Dunsavage wrote: I'm confused. Can someone sum this up for me? I'm from USA. Last Thursday, the German parliament passed a concept that had been proposed by the German government parties. The concept concerns the German criminal code and is meant to tighten laws against cybercrime. According to the new §202c, anybody who prepares a crime by building, supplying, distributing or making available passwords or security codes for data access or typical computer programs whose purpose is to prepare or commit such a crime, can be fined or sent to jail for up to one year. There were other changes concerning §202, but the one mentioned above is the one most criticized. Many people say that it's not possible to distinguish between programs that might be used to prepare a crime and programs that serve to detect vulnerabilities and secure computer systems (I personally agree with that statement). Therefore, §202c could criminalize many tools that are frequently used these days, for instance port scanners etc. There is no clear definition given in §202c and at the end of the day a German court might have to decide in individual cases. The intention of §202c, however, seems to be to criminalize only software that might cause a damage. In order to become a law, the concept has to pass the German Bundesrat (upper house of the German parliament) as well. This could happen in July. Then the new concept would become a law shortly thereafter. It could affect openSUSE (in Germany) since the distribution of programs that fall into above mentioned category (yet to be clearly defined) is then forbidden. HTH, Th. PS: This topic is not of technical nature and should be discussed on opensuse-project. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Undelete
Randall R Schulz wrote: [...] People have been known to replace the rm command with a script or shell procedure that moves the target files to a trash directory instead of deleting them outright. It still doesn't help with, e.g, and application's Save as... function that overwrites an existing file, but it's something. That's why you should have a backup of important files. If the file is deleted or the disk crashes or whatever and a user does not have a backup, then the user can only blame himself for this trouble. With all these large external USB disks available nowadays, it shouldn't be a problem to make regular backups even for home systems. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] OpenSUSE locks up
Randall R Schulz wrote: [...] (Note that while the S in S.M.A.R.T. is technically SCSI, many SATA and (I think) IDE drives support it, as well). S.M.A.R.T. means Self-Monitoring, Analysis and Reporting Technology and the S has nothing to do with SCSI (although this technology has been introduced for SCSI disks first before it became pretty much a standard for all SCSI and ATA disks in the mid 90ies). -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Small database program?
Nick Zentena wrote: Is there something small and simple? All I want is to set up a simple home database. Items,serial numbers maybe a couple more fields. Every time I go looking all I seem to find is stuff geared at running a Fortune 500 company. sqlite3 (+ GUI frontend if required) Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] how to burn my cpu?
Eberhard Roloff wrote: [...] I am looking for an app that makes my CPU as hot as can be. Compile the Linux kernel in an infinite loop with make's -j option for multiple jobs. Or use gcc source instead of the Linux kernel. In principle, this should test your CPU and RAM quite well. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Strange behaviour of ACPI on laptop
torben wrote: [...] Secondly, I would prefer to use the old kernel configuration as a starting point for the configuration of the new kernel. Can I somehow export the configuration of the old kernel and use it as a starting point for the new kernel? Yes. With SuSE kernel sources, you can just run make cloneconfig in the new kernel source tree and this will clone the configuration of the currently running kernel. With vanilla kernel sources (however, it works of course also with SuSE kernel sources), you can copy the existing old configuration (e.g. /boot/config-2.6.xyz or /usr/src/oldkernel/.config) to a file called .config in the new kernel source tree and then run make oldconfig. Both approaches do not use a build directory, but you might not want to use one anyway... HTH, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] time command
Herbert Georg wrote: [...] I'm trying to use the command time to access information about a job, but without success. The man page tells me that I can use options to time, like: time -o output-file command command-args time -v command command-args time --help but nothing of this works. All I get is (in the last example): bash: --help: command not found real0m0.001s user0m0.000s sys 0m0.000s I get the same thing in all option cases. Why doesn't it accept any option, and keep interpreting the options as the commands? You're using the shell builtin time command which does not support the options. Try /usr/bin/time and it will work. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] uninstall source
lists Guillot wrote: [...] I often install packages from source with ./configure, make, make install. Sometimes I'm not happy with the software for whatever reason and I want to get rid of it. I don't actually know how to do this, till now I've always done make clean or make distclean if available, and then deleted the sources. This procedure does *not* uninstall the software. But I installed something the other day that messed up something else, and uninstalling in this manner did not fix the problem. So how does one really fully undo what is done by ./configure, make, make install? Unless the Makefile provides an uninstall target, there is no way to uninstall software that has been installed with make install. On a system based on RPM package format, you should only install RPM packages and nothing else. Otherwise, the consistency of your RPM database is no longer guaranteed. [...] So I uninstalled iconv as mentioned above, but the problem didn't go away. to my surprise I found that /usr/local/bin/iconv was still there. If I delete it the problem is solved, so that's ok. The question is that I thought make clean would have got rid of it, and now I wonder what other junk is left lying around from other source packages I've installed and removed. make clean or make distclean cleans up the source/build directory. These Makefile targets do not remove anything that has already been installed. How does one really uninstall? How do you know if everything's been removed? You should only use RPM packages - these packages can be installed and uninstalled without problems. Either get a ready-to-use RPM package from a repository, or build an RPM package on your system, or (at least) use checkinstall to build an RPM package (instead of using make install). Anything else will sooner or later screw up your system. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] uninstall source
[EMAIL PROTECTED], please stop sending private emails and reply to the list only. Furthermore, please avoid top-posting. [EMAIL PROTECTED] wrote: if u try to ./configure make. make install (then don't happy with this installation) how abt this... don't delete your configured source (after your ./configure ... make... make install) then do this make uninstall the good packages should provide 2 way, how to make it installed n how to make it uninstalled..(am i right??) 1.) Most packages do not provide an uninstall Makefile target. 2.) Even if a Makefile provides an uninstall target, it does not solve the inconsistent RPM database problem. On an RPM-based system, you have to use RPM packages in order to keep your system in a clean state. Anything else should only be used by experts who really know what they are doing... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] building RPMs
lists Guillot wrote: [...] You know, I've always been uncomfortable with building from source for precisely the reasons you say, but it never occurred to me to build my own RPMs. I somehow had an idea that it would be a complicated process that only people who *distribute* software (and not all of them!) would find worth doing, and that for your home system you just had to go with make install. The make install procedure has some obvious drawbacks you might have just observed: uninstalling the software is a problem, the consistency of your system (RPM database) is a problem, etc. When you install software via make install, then the RPM package manager does not know about it and cannot resolve dependencies or tell you about conflicts with already installed software. Therefore, you're better off installing all your software as RPM packages (if possible). I'm going to follow your advice. Johannes intimated above that checkinstall has a bad reputation, and you also don't seem to think too highly of it with your (at least). So what would be a good tool for this? I see that I have rpmbuild on my system so I'm tempted to use that, though the man page is not exactly a detailed instructions manual. checkinstall is a useful tool but has also some drawbacks. I think just give it a try. I would recommend the following order for installing new software (note that this is my personal opinion): 1) Think twice whether you need a new version or new software. New software will not only have new features, but most likely also new bugs. 2) Try and find a ready-to-use RPM with the version you like and built for your system (e.g. SuSE 10.0). You should be able to install this RPM without problems. If the RPM has lots of dependencies, it might be easier to add the corresponding repository as installation source (if a repository is available). 3) If a ready-to-use RPM does not exist, try to get a source-RPM (SRPM) and rebuild it on your system using rpmbuild. This will create an RPM package on your local system which can then be installed. 4) If neither 2 nor 3 is possible, compile the source code on your local system and a) as an expert create your on .spec file and an RPM package or b) as a normal user use checkinstall to create an RPM package. This RPM package can then be installed as usual. You could also borrow a .spec file from an older SuSE SRPM and adjust it according to your needs (and for the new version). 5) As a last fallback solution, you can use make install. However, think twice and note the obvious problems it might cause. If you want to learn more about RPM, have a look at the Maximum RPM book: http://www.rpm.org/max-rpm/ - however, it might be a bit tough since it is targeted rather at the developer than a normal user. The checkinstall homepage can be found here: http://asic-linux.com.mx/~izto/checkinstall/ - but I am sure you also find a SuSE repository that offers checkinstall RPMs. Hope this helps. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] vote for the 10.3's slogan
Alexey Eremenko wrote: openSUSE - Linux for open minds has WON ! Very nice indeed. The vote ends Friday, 27th April. Today is Saturday, 21st April. You got it? I miss the option I don't like any of those slogans - I think it would have a good chance to win... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] *help* Built 2.6.20 kernel - boots fine, no keyboard, no mouse, no network
david rankin wrote: From: Thomas Hertweck [... create link linux-linux-2.6.20 ...] That should not be necessary. Sorry, everything I have read, said do it, so I just did it In principle, the link /usr/src/linux is not required at all and has mainly historical reasons. It's nowadays only used for convenience. [...] The machine boots just fine on 2.6.20.6 and 2.6.20.7, the only problems are (1) keyboard; (2) mouse; and (3) network. Everything else works fine (Apache, MySQL, smb, nmb, sata, etc...) Well, you cloned the old configuration, so all that stuff that worked in the old kernel should be available in the new kernel as well. However, it does not work at the moment, so there must be a problem which indeed could be related o udev 081 # udevinfo -V ...to udev. udev is responsible for creating and removing device nodes, renaming network interfaces, etc. If devices were missing, then this would of course explain why some of your components (mouse, etc.) do not work. However, other components seem to work, so it's not failing completely or in a consistent way which does not allow any accurate conclusion. All your problematic components are hardware-related, so I am not surprised that apache, mysql etc. work as expected - that's all pure software, isn't it? Out of the entire list, the only requirement I did not update was udev because the documentation said it was not a good idea. Well, udev is a core component and I would agree that messing around with udev is not such a good idea ;-) I am not really sure what advice to give in this situation, maybe others can contribute their point of view in this discussion. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 2.6.20 on 10.2 (again)
John Andersen wrote: [...] So what's your point? Did you see me say it was part of the kernel? I raised the question because your statement Beagle [...] and something else [...] were missing from 2.6.21, Msg-ID [EMAIL PROTECTED], did indeed sound to me as if you were trying to say that Beagle is part of the kernel and it's now missing in the latest KOTD. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Call for Testing: openSUSE 10.2 Bootloader Test
John Andersen wrote: [...] Well that brings it full circle then, doesn't it Thomas? The OP posted that he DID try the new kernel (the subject of this thread) and it did NOT work for him, and others (mistakenly) suggested he modprobe it, only to be told by you that its in the kernel and he need do nothing of the sort. Its all starting to sound a little like a Monty Python skit isn't it? The OP posted in [EMAIL PROTECTED] that it worked in a similar way I've described in my email - USB_DEVFS is enabled, all that had to be done is mounting the filesystem. So what's your point here? Just trying to bash other people? PS: This is a rhetorical question, so no answer required. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] *help* Built 2.6.20 kernel - boots fine, no keyboard, no mouse, no network
david rankin wrote: This is really comical - sort of. I am running opensuse 10. Is that 10.0? I have been through heck compiling a new kernel. I got the 2.6.20 kernel and did the following Get the latest and greatest kernel, unpack it into /usr/src/linux-2.6.20, create a symlink linux-linux-2.6.20 That should not be necessary. cd linux zcat /proc/config.gz config make oldconfig make make modules_install make install. edit grub to boot from new and old kernels. Did make install create an initrd etc.? [...] Anybody want to take a stab at where I screwed up?? Well, did you read Documentation/Changes in the kernel source tree? Does your system meet the Current Minimal Requirements? If it's indeed a 10.0 system, then I have some doubts... Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Call for Testing: openSUSE 10.2 Bootloader Test
STDIN wrote: [...] I tried, its not there: # modprobe usbfs FATAL: Module usbfs not found. As already mentioned several times on this list: usbfs is not a self-contained kernel module. It's part of usbcore which is most likely compiled as a module. Therefore, you should get your USB_DEVICEFS with usbcore.ko if this feature had been enabled in .config at compile time (which is true for kernels to be included in 10.3 and the latest patched kernel for 10.2). All you need to do is mount the filesystem, or change /etc/fstab to do it automatically when the system starts up. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Hyperthreading
Randall R Schulz wrote: [...] HT was better than a non-HT single-core CPU, [...] Unfortunately, there are applications that slow down when HT is enabled. Therefore, it's highly application-dependent whether a HT-enabled single-core CPU is better than a non-HT single-core CPU. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] kernel updates
Marcus Meissner wrote: [...] All other distros do the same as far as I know. Care to give an example where this is not the case? My Fedora system does not remove old kernels when a kernel update is installed via yum, the default software management tool. It just adds a new entry in the boot manager configuration file which then becomes the default boot option. The same holds for kernel sources, etc. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] 2.6.20 on 10.2 (again)
John Andersen wrote: [...kernel 2.6.21...] Even if it is stable, there are some packages missing which makes it unusable for some things. Beagle (no great loss imho) was missing and something else, which escapes me (but was a show stopper) were missing from 2.6.21. Beagle is part of the kernel?? I thought that's a desktop search engine based on *argh* Mono... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] the new openSUSE kernel and the IDE - A different slant
scsijon wrote: [...] I have scripts and software that create drives with links to those in other systems as part of the security and backup routines I use, some are to scsi drives, some are ide and some sata, the drive type in most cases relates to the function performed on the site or spanning sites Before I can consider going to 10.3 :- I shall have to scan, amend, rewrite, binfile at least 20 scripts for cross connection, another 38 for backup and seven security plus workout how I get around the drive type and now limited drive identity\numbers problem and that's just the linux side, how my mini's deal with it I am yet to imagine, thankfully I don't have any mainframes any more. I also suspect it means some ten to twelve thousand dollars additional hardware needed per site, say a quarter of a million dollars plus extra at the four centres and spares of course. [...] Frankly, I would consider your backup strategy as broken if it relies on a particular hardware being available as hd? instead of sd?. Of course, it's difficult to say without knowing the details, but it sounds rather strange. Maybe your backup etc. strategy is designed in a suboptimal way. What would happen if your backup server broke down and you had to switch to another one with different hardware? Would you have to change all your backup scripts, etc.? That should not happen. If we have to change our backup servers in the company, it's a matter of seconds not at least 200 hours of work ahead... CU, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to install 2.6.20 kernel on opensuse 10
Kenneth Schneider wrote: On Wed, 2007-04-11 at 14:59 -0700, David Brodbeck wrote: david rankin wrote: Will my .config from the 2.6.13 kernel source I got from yast work with the new kernel source from kernel.org? ^ Yes. Copy the old .config file to the new kernel source directory, then do 'make oldconfig'. It will use the old settings where it can and ask you about any new ones. No need to copy the file, just use make cloneconfig instead (without the quotes of course). No. make cloneconfig ist a SuSE extension and won't work with vanilla kernel sources from kernel.org. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to install 2.6.20 kernel on opensuse 10
david rankin wrote: [...] What am I doing wrong? I can't get make cloneconfig to work. Please have a look at my other email in this thread. make cloneconfig does not work with vanilla kernel sources. Before you continue to play with the new kernel, please have a look at Documentation/Changes in the kernel source tree. On older systems, it's likely that you have to upgrade other packages as well (e.g. udev), not just the kernel. You will find the Current Minimal Requirement in the file mentioned above. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Re: Please make available a 2 DVD set for openSUSE 10.3
Alexey Eremenko wrote: I'm happy, that at least one man around is supporting me ! I support this as well. All this repository stuff is getting out of hand from my point of view, so having at least all the official SuSE packages on a nice (modular) DVD ISO set is something very useful... Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Another grep question
James Hatridge wrote: I'm not sure if this is a grep or awk question. I have a file with many records of 4 lines each in it. Like this: 1 2 3 4 1 2 3 4 etc, etc I need the file to look like this: 1 tab 2 tab 3 tab 4 1 tab 2 tab 3 tab 4 etc etc How can I do this? $ paste - - - - file.in file.out Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] udev problem
M. Todd Smith wrote: [...] In /etc/udev/rules.d/50-udev-default.rules, there is a line that looks like: KERNEL==nvidia*|nvidiactl*, GROUP=video Reading through the man for udev rules tells me that I *should* be able to add a line like this: KERNEL==nvidia*|nvidiactl*, GROUP=video, MODE=0666 ^^^ quotation mark missing here I think it should read something like: KERNEL==nvidia*|nvidiactl*, NAME=%k, GROUP=video, MODE=666 Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel Building Resources Recommendations
Gaël Lams wrote: [...] I will have to rebuild kernel to be used in VMWare guests for various servers (and rebuild them after each kernel security updates), I'm interesting in understanding and testing the whole procedure. As I wrote in my last email, there are several ways to deal with this problem. If you have to support many systems and not just your local desktop system, then the simplest way might be an RPM package that replaces the default usbcore.ko (from the SuSE kernel installation) with a new (your own) version that has USB_DEVICEFS enabled. This might minimize the possible side-effects as only a single file is changed. [...] I used to speak german. Would it be possible to have a link to this how-to? Maybe we could write something to be put on the OpenSUSE wiki. This project started a long time ago when there were many questions on suse-linux (the German mailing list, now opensuse-de) concerning the SuSE kernel, kernel patches, recompiling a kernel, etc. As the official German kernel howto was (and still is: Linux Kernel HOWTO v0.80-3, 1. Juli 1998) out of date, I decided to write my own howto... I could have written this howto in English - however, many people on suse-linux mentioned that they have problems understanding technical literature in English, so I decided to write it in German (I think, there is quite a lot of kernel-related information available in English, at least concerning the Vanilla kernel; and there is the nice document WORKING WITH THE SUSE 2.6.x KERNEL SOURCES written by Andreas Gruenbacher, http://www.suse.de/~agruen/kernel-doc/, also part of the SuSE kernel sources: README.SUSE). Luigi di Lazzaro translated my little howto into Italian. To cut a long story short, here it is: http://www.thomashertweck.de/kernel26.html Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel Building Resources Recommendations
Patrick Shanahan wrote: [...] 16.55 wahoo:~ rpm -q --changelog kernel-default | head (none)* Fri Mar 09 2007 [EMAIL PROTECTED] - Enable CONFIG_USB_DEVICEFS (#210899 and a zillion others.) Turns out that vmware isn't going to change anything, so making our users (and executives) have to build their own kernels is not something we should be doing. I was wrong, sorry. Sounds good. The next task then is to convince the SuSE guys to ship NVIDIA kernel modules... *duck and run* ;-) -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel Building Resources Recommendations
Patrick Shanahan wrote: * Thomas Hertweck [EMAIL PROTECTED] [03-12-07 17:12]: Sounds good. The next task then is to convince the SuSE guys to ship NVIDIA kernel modules... *duck and run* ;-) but they are not open-source. Illegal. It was a joke, ok. Did you see the email address in the kernel changelog? That might tell you something. If not, then you won't understand the story... OK, time to end this thread, problem is solved and there is no need to continue this chat. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] halt / reboot waiting time
Anders Norrbring wrote: I have a small problem, when the system goes for shutdown or reboot, the waiting time is too short for a process. I run a VMware on the system, it can take up to 7-10 minutes to shut down all VMs properly, and the system is too eager to shut down. So, before VMware has finished its shutdown properly, SUSE kills the processes which results in heavily screwed up disks in the VMs. How can I extend the waiting time for shutdown/reboot so that VMware can exit properly? man shutdown: -t sec Tell init(8) to wait sec seconds between sending processes the warning and the kill signal, before changing to another runlevel. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] [Repost]: Using an extra driver at install-time
Philippe Andersson wrote: Can anyone out there please help ? Any pointer to documentation explaining how to hack / customize the OpenSuSE installer would be welcome. I am not sure whether this is of any help in your current situation, but it might be worth a look: http://ftp.suse.com/pub/people/hvogel/Update-Media-HOWTO/ In old days, one could use a floppy disk to provide additional modules at boot or installation time, respectively. You pointed out that your system has no floppy drive - maybe SuSE has meanwhile implemented other ways of providing additional modules. Unfortunately, I cannot answer this question since I am still using an old version ;-) Maybe just boot the installation medium and check what methods are available... Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] halt / reboot waiting time
Anders Norrbring wrote: [...] Well.. Not really. I want the scripts to actually WAIT until VMware has exited. The -t option to shutdown is used to set the time the system waits from sending the shutdown message to screen until it starts the shutdown process. No. It should be the time between sending a SIGTERM and a SIGKILL signal. The time when to start the shutdown process is given as a shutdown argument, not an option. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] halt / reboot waiting time
Anders Norrbring wrote: Thomas Hertweck skrev: [...] No. It should be the time between sending a SIGTERM and a SIGKILL signal. The time when to start the shutdown process is given as a shutdown argument, not an option. My mistake, sorry. Anyway, what's the default time? I can't see any option to find out.. You should really familiarise yourself with the manual pages... :) man init: When init is requested to change the runlevel, it sends the warning signal SIGTERM to all processes that are undefined in the new runlevel. It then waits 5 seconds before forcibly terminating these processes via the SIGKILL signal.[...] telinit can also tell init how long it should wait between sending processes the SIGTERM and SIGKILL signals. The default is 5 seconds, but this can be changed with the -t sec option. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel Building Resources Recommendations
Randall R Schulz wrote: [...] 1) Specific recommendations on the most expedient way to endow the kernel in openSUSE 10.2 with this USB support. In particular, is it possible to simply load (or build and load) a kernel module to reenable the USB filesystem (known variously as usbfs, usbdevfs, USB_DEVICEFS, or /proc/bus/usb)? usbfs is not implemented as a separate kernel module. As far as I know, you need to recompile the usbcore module if you want to re-enable usbfs. 2) Assuming a full kernel rebuild is required, I'd like to get some recommendations for informational resources of a tutorial nature on how to build a late-model kernel with emphasis on anything openSUSE-specific or -related. I found this How-To Forge article: http://www.howtoforge.com/kernel_compilation_suse. If anyone is familiar with it and can comment on its usefulness, that would be helpful. Well, I would not really recommend this Howto, but that's just my personal opinion :) [...] 1. Install kernel-sources and kernel-syms 2. copy .config and Modules.symvers from /usr/src/linux-obj/your kernel to /usr/src/linux 3. make oldconfig 4. make menuconfig, select the usbfs 5. make modules make modules_install 6. change noauto to auto for the usbfs in fstab 7. reboot 8. ta-da I think, the following procedure might work (however, it's untested since I do not have a 10.2 installation): $ cd /usr/src/linux $ make cloneconfig $ vi .config change # USB_DEVICEFS is not set to USB_DEVICEFS=y $ make oldconfig $ make drivers/usb/core/usbcore.ko and then install this new module (i.e. replace the standard usbcore.ko module in /lib/modules/`uname -r`/ with this new one). That should avoid the time-consuming process of compiling all modules (make modules). However, this will only work if all the USB_DEVICEFS functionality is really contained in the usbcore module (I think it is but I am not 100% sure). If not, then you have to re-compile additional modules and a make modules might be a safe way to do it. Maybe someone using 10.2 can report which way works best. Please note that this method is not using a build directory, i.e. it will build the module(s) in place (i.e. in the kernel source tree). Also note that you'll have to go through this procedure again after upgrading a SuSE kernel. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Kernel Building Resources Recommendations
Randall R Schulz wrote: [...] It did not seem to lead me totally astray. I went with the suggestion of creating an RPM (because I was using one computer to do the build but the result was intended primarily for my friend's computer). This is where things fell down, probably because I don't know enough about how kernel RPMs work, what happens when they're installed. How the source RPM is fabricated (in particular, why I can have more than one kernel but apparently only one set of kernel sources installed at any one time). A make rpm creates (AFAIK) only a very basic RPM package. If you really need to build a proper RPM package for your SuSE system, then I recommend using the SuSE Kernel Source src.rpm and making changes therein. If you know of better resources for newbie kernel builders, I'd love to hear about them. I wrote a kernel howto (actually this project started in 2002) which I think is well known among German-speaking Linux users - I get quite a lot of positive feedback. However, since it's written in German it might not be particularly helpful for you... [...some suggestions...] I think it's too minimal for practical purposes, especially if you want to transport the new kernel to another system. True. I thought you only wanted to change your local machine. [...] When I ran menuconfig I gave the configuration a variant name, but somehow I still didn't get a new build directory. I think this is something different. A build directory is where you build the actual object files, modules, etc. and the final kernel image. In principle, /usr/src/linux-obj (or a subdirectory, respectively) is such a build directory - when using a build directory, you separate all the files created during a build from the source tree (the kernel source tree is always a clean tree). If I understood you correctly, then you wanted to give your new kernel a unique kernel release (which in general is a good idea when compiling your own kernel). This would translate into a new /lib/modules/ subdirectory. Did you mean such a directory? Or really a build directory? From the info I've seen so far, it's difficult to say what might be wrong with your approach and why the vmware script fails on the other machine. I think, there are several approaches to tackle the problem: a) build your own complete kernel with USB_DEVICEFS enabled (via make rpm etc.) b) build your own complete kernel via src.rpm (default config adjusted) c) only build a new usbcore.ko module for the standard SuSE kernel, copy it or package it as RPM, and install it on the other machine(s) d) ... From my point of view, c) might be the quickest and easiest solution as it does not require to install a complete new kernel, kernel sources, etc. on the other machine (remember, if you install a completely new kernel, you or the RPM package also need to take care of the initrd, bootloader configuration, etc.). You would only replace a single kernel module in the SuSE standard installation. However, it seems as if you've already managed to install and boot a new kernel and it's now only the vmware part that's going wrong. Correct? Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Fortran 90
James Knott wrote: [...] So back in those days, you'd have computers languages designed for business or scientific use. It wasn't until the early - mid sixties that general purpose computers started to appear. With computer power so cheap these days, such optimization is no longer needed. This is not entirely true. Just look at what is nowadays known as reconfigurable hardware or FPGAs. This is a form of adjusting hardware in order to run software (algorithms) with optimized speed. In other words, such an optimization still exists and is used in HPC. Sometimes, it's even necessary because the compute requirements easily exceed traditional general purpose processor solutions... -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Fortran 90
Kai Ponte wrote: [...] People still use FORTRAN? Your knowledge about programming languages seems to be quite limited. For instance, Fortran is used in scientific and high-performance (numerical) computing, and many companies still have old Fortran code to maintain. Therefore the answer is: yes, people still use Fortran. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Is opensuse@opensuse.org == suse-linux-e@suse.com ??
Dave Howorth wrote: Robert Lewis wrote: Thanks, that caused me to go the extra mile. It turns out that that I had so many headers that the X-Mailinglist one was off the screen at the bottom. To make matters worse I couldn't expand the screen to catch all the headers. Furthermore, Thunderbird doesn't put up a scroll bar when the headers exceed the screen size. I see the same symptoms but fortunately my screen is big enough that the header is visible :) Thunderbird does take space from the message list area to expand the message/header display area, but there are more headers that I can't see and there's no scroll bar or other way to scroll that I can find. https://addons.mozilla.org/thunderbird/1003/ TH. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] vmware and kernel version problem
I hope some day people will realize that UTS_RELEASE is nowadays defined in the header utsrelease.h (and no longer in version.h) and that SUSE uses a build directory, i.e. generated header files like version.h are *not* in the source tree (/usr/src/linux/include/linux) but in the build directory (/usr/src/linux-obj/arch/flavor/include/linux). The link /lib/modules/`uname -r`/build should point to the build directory. Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] fsck running amok
Darryl Gregorash wrote: On 2007-01-22 13:09, Thomas Hertweck wrote: snip I don't know details about ReiserFS - this FS has been banned from all our systems a long time ago. I know a bit about ext3 and xfs though. Interesting. I'd be interested in knowing why, if you can share that information. We have used ReiserFS when it became known and when it was available in Linux distributions. However, at that time we faced lots of weird problems that were caused by the filesystem (yes, these were problems with the FS and not the hardware) and reiserfsck was not working very well, i.e. if something went wrong, we had situation where reiserfsck made things much worse instead of repairing the FS. We then decided that this is certainly not a filesystem we would like to use in a production environment. Since we usually have to deal with large systems (tens or hundreds of TiB) and (very) large files, we decided to try xfs (which we already knew from our SGI servers) - xfs was designed and optimized for large filesystems and large files right from the beginning. It turned out to be a very good decision and, thus, there was no need to come back and try ReiserFS. On thin clients or laptops and/or dual-boot machines, we are using ext3. It's a bit safer when it comes to losing data at power cuts (which is due to the way the journalling works) - our desktops have no UPS. Furthermore, there exist tools to access ext partitions from Windows OS. When the filesystem is marked as clean, then there is usually no need to do an fsck or to replay the journal [...] I thought you had to run fsck to determine if the filesystem is clean? Sorry, that was misleading: when the filesystem is marked as clean, then there is no need to run an entire filesystem check/recovery over the whole partition or to replay the journal. That might be a better phrasing. I hope you know what I mean. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] fsck running amok
J Sloan wrote: [...] Sorry, but that just sounds too bizarre to me - we've been using reiserfs in our data center for years, since it's the default filesystem on suse linux. It's been rock solid here, and is also the fastest journaling filesystem we've found. Sorry, I am not interested in a thread about filesystems. Darryl asked the question and I answered it. If ReiserFS works for you very well, that's fine, I have no problems with that statement. We have done a lot of testing quite some time ago and made bad experiences which is the reason why we decided against ReiserFS - calling that bizarre just because it doesn't conform to your own experiences is not very nice. In our situation with big files and big filesystems, xfs does now a very good job, so why should we change? I understand that reiserfs maintenance going forward may slow down, so we'll have to eventually settle on a different filesystem, but there's no reason to suddenly change all of our stable systems on a political whim. Nobody said you should - where did you get that idea from? Darryl asked why we don't use ReiserFS and I answered exactly that question. I don't really care what you do and what filesystem you use - everybody has the choice and that's a good thing. If you're happy with ReiserFS, fine! We're happy with xfs. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] fsck running amok
Darryl Gregorash wrote: [...] The reason I wrote that way is because my boot.msg contains the following: Checking file systems... fsck 1.36 (05-Feb-2005) Reiserfs super block in block 16 on 0x349 of format 3.6 with standard journal Blocks (total/free): 2008112/1231671 by 4096 bytes Filesystem is clean Replaying journal.. etc Am I correct in viewing the whole thing, super block plus tree structure, as the file system? It would seem that first there is a check for a valid super block, then the tree is checked for consistency, and only then is the journal replayed. (Other journalling f ilesystems may vary in specifics, of course.) I don't know details about ReiserFS - this FS has been banned from all our systems a long time ago. I know a bit about ext3 and xfs though. When the filesystem is marked as clean, then there is usually no need to do an fsck or to replay the journal - the filesystem should be in a consistent state and it can be mounted (there will be a forced fsck if max mount count has been reached etc. or if the fsck is forced by the user, e.g. via /forcefsck). If the filesystem is not marked as clean, then the journal is replayed in order to bring the FS into a consistent state. If it works, then the FS is marked as clean and will be mounted. If something goes wrong, then a complete fsck might be necessary. If it's the root filesystem, then you might end up at a root console and the system asks you to perform an fsck manually. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Is there an RPM command that will do this?
Kai Ponte wrote: [...] That's because man pages are less than useless. Nonsense. Manual pages are perfect as a reference, nobody really knows all the options for Unix commands. They are at least 10 times better than info pages. However, both are usually not meant to be a tutorial or a beginner's guide. Read the maximum RPM handbook[*] if you need a full-blown introduction to RPM and not just a reference. [*]http://www.redhat.com/docs/books/max-rpm/ -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] fsck running amok
Darryl Gregorash wrote: [...] I may have mentioned in another note that this is likely due to the introduction of journalled file systems. I believe the situation is roughly this: Re-running the journal on boot allows the system to keep the file system in better order. Before that happens, however, a consistency check must be done on the file system. Journal replaying itself is done to ensure the consistency of the filesystem. Since the journal contains a chronological log of recent metadata changes, it's able to simply check the portions that have recently been modified - which is a matter of seconds. It does not re-order anything. After this step, the filesystem is usually marked as clean (unless something goes wrong and an fsck might be forced). -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Recommended Website Tools
I use bluefish. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Moderator - Reply to settings
Cristian Rodriguez R. wrote: [...] This behaviuor is incredible annoying No, it's not - I like it. It's maybe annoying for you, but you can't speak for everyone! -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Asus M2N SLI
BandiPat wrote: [...] Nobody has to bait you Ran, you manage to slip the noose around your head all by yourself. We watch, we read, we laugh at you falling face first into the mud you make for others. You never seem to learn though and really with so many failures on your part, it would seem some knowledge would be gained along the way. Frankly, I am laughing at you, not at Randall. Remember the old thread on opensuse? http://lists.opensuse.org/opensuse/2006-10/msg00352.html You've already tried Asus-bashing at that time - it's a good sign that many people don't believe you and oppose your position! Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Which flavor of openSUSE for Intel Core Duo chipset
Randall R Schulz wrote: On Friday 29 December 2006 18:53, John Andersen wrote: [...] It will work with either i386 or X86_64, but in my tests its faster using x86_64. (aka EMT64). What are those tests? My on-line research indicates that there are but a very few circumstances in which the Core 2 Duo chips are faster when used in their 64-bit mode. I have no experience with Intel Core 2 Duo chips yet, but with Intel Xeon server CPUs which allow to use 32-bit or 64-bit OS and/or software. Our experience is that we get (same hardware specs; same software) up to xx% better performance using x86_64, i.e. we've made similar experiences to John. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Which flavor of openSUSE for Intel Core Duo chipset
Randall R Schulz wrote: [...] What sort of an application mix do you run? We develop and run software for seismic data processing on our Linux clusters (some 64-bit, some 32-bit). In other words, it's mainly number crunching. We saw performance differences of up to 30% on the same hardware. We asked Intel about this unexpected behaviour but they've not been able to come up with an explanation yet, it's still under investigation. I agree with you that Xeon's and Core 2 Duo's characteristics might differ, but nevertheless it's an interesting observation. Unfortunately, I can't give any more details at the moment. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Diff. between OpenSuSE and SuSE Linux Enterprise
Brett I. Holcomb wrote: [...] Fedora is really an experimental test distro for RHEL so comparing OpenSuse with it is not really accurate since OpenSuse releases are stable, production ready. You're kidding, right? -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Diff. between OpenSuSE and SuSE Linux Enterprise
Brett I. Holcomb wrote: On Wednesday 27 December 2006 15:54, Thomas Hertweck wrote: Brett I. Holcomb wrote: [...] Fedora is really an experimental test distro for RHEL so comparing OpenSuse with it is not really accurate since OpenSuse releases are stable, production ready. You're kidding, right? No. My understanding and the explanation I was given is that OpenSuse releases - not the betas - are much more stable than Fedora Core is. Or in other words: you have no experiences with FC systems and (as it seems) with production systems in general but you make the statement that opensuse releases are much more stable than FC releases and opensuse releases are ready for production. Did I get this right? Well, although I like opensuse I think reality has significantly differed from your statement in 2006... Cheers, Th. PS: Brett, please read http://www.river.com/users/share/etiquette/edit.html before you reply to any other email on this list. No top-posting, please. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Ganglia 3.0.3 RPM ?
cifroes wrote: Is there any rpm for opensuse 10.1 for ganglia 3.0.3? Or is there any easy way to install it? the rpm included is from a very old version of ganglia... http://repos.opensuse.org/home:/ganglia/ 10.1 directory is empty, but you should be able to use the src.rpm from e.g. SUSE_Factory/src and rebuild the rpms on your local system with rpmbuild --rebuild ganglia-monitor-core-3.0.3-23.6.src.rpm. This should result in five rpms: ganglia-monitor-core, ganglia-monitor-core-devel, ganglia-monitor-core-gmetad, ganglia-monitor-core-gmond, and ganglia-webfrontend which can then be installed as usual. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to build a cluster with 64 bit Opterons and SUSE?
HG, I think you should have a look at OSCAR [*]. I'll send you another email as PM. Cheers, Th. [*] http://oscar.openclustergroup.org/ -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] How to build a cluster with 64 bit Opterons and SUSE?
HG wrote: [...] We would like to build a cluster of multiple Opteron based computers. The usage is for scientific computing (multiple scientists running large processes usually at the same time... :-) Mainly computing but also some large flat file access, no DB stuff. [...] How? Can't find anything really about getting any SUSE version or 2.6 kernel for that matter to work in cluster. Anybody have any experience? Any pointers to resources? Opensourse? Commercial? Whatever? Building a proper cluster and cluster environment is not as easy as it might sound and involves a bit more than just buying some PC boxes and network cables, but I am sure you're aware of this problem ;-) First of all, I think you need to exactly identify the purpose of your cluster. From your description, it sounds like you're going to build a compute cluster for mainly CPU-bound processes. Think about the size of your cluster, the RAM per cluster node (which depends on the type of applications and, of course, your budget), and the connection between cluster nodes. Think about the local disk storage per cluster node you might require, the remote disk storage, and so on. And don't forget the cooling solution and power supply... I would recommend buying a properly configured cluster solution with some decent cluster management software. Building a big cluster on your own is a tedious task. If it's a small cluster, it might be much easier. Well, all of this somehow depends on your budget. You also need a good queuing system which allows to assign different properties to cluster nodes (OSS solutions available). There are also pretty good cluster monitors around (e.g. http://ganglia.sourceforge.net). We're using several HPC clusters with thousands of CPUs, all running Linux OS, Kernel 2.4 or 2.6 (I must admit it's RHE not SuSE but this should not really matter in this context). Sorry, I can't give any more information in public. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Finding which version of a driver is installed?
John Bright wrote: Try checking man modprobe,insmod rmmod to learn about linux modules. also check /etc/modprode.conf for the modules loaded on boot. kernel modules listed in /etc/modprobe.conf aren't loaded on boot. CU, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] On my SUSE linux 10.1 kernel: ide: failed opcode was 0xef
Boyd Lynn Gerber wrote: On my laptop I just started getting this error below about 3 weeks ago. I have tried to track it down, but I have been unsuccessfule. The message in the log is... Dec 6 06:33:35 blg kernel: ide: failed opcode was: 0xef This is the last message in logs and then I am not able to close any windows in KDE or write to the HD. The only way to get things to work is a power off. I have been unable to track down what is causing this problem. It can happen from an hour to 6 days. That is the longest I have made it. I have another laptop which is almost exactly the same and it never happens. Any ideas on how to find the cause? I do notice that the machine near the right side of the keyboard some times is a bit warm, but I am not sure if that is the problem. If the machine is hung and I go to one of my X-terms on the machine I quicly find out that it is hung. An other symtom is the clock in the lower right conner stops advancing. Also the repeat on the arrow keys stops functioning. It might be completely unrelated, but you never know... I have similar problems on one of my laptops running SuSE 10.0. The problems started some weeks ago. Sometimes, I recognise that the load starts to increase (xosview is running all the time) - from this point on, I am no longer able to close windows or switch to other desktops. The mouse is still working. If the focus is on a terminal window, I can still enter a command and press return but nothing happens. The load is increasing to about 25 and then more or less everything stops working (keyboard no longer responding, etc.). There is no information in /var/log/messages. I've checked memory with memtest86 over night: no problems. I've check the disk with the manufacturer's disk tool (it's a WD disk): no problems. I haven't been able to figure out the trigger for those problems. It might happen any time. The only solution is a power off. Other laptops here (also SuSE 10.0) work without such problems... Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] NO PERSONAL REPLIES TO POSTINGS ON THIS LIST!!
Hoper Edei Deixai wrote: [...] Not Thunderbird nor any webmail client. Please don't start this discussion again. Thunderbird has an extension which provides a list-reply functionality. See the November archive of this list for details (threads named Seamonkey 1.0.6 rpm? and Thunderbird reply to list feature). Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] NO PERSONAL REPLIES TO POSTINGS ON THIS LIST!!
John, answering private emails on a mailing list is considered bad behaviour!! Please don't do it! There was a reason why I sent my previous email only to you and not to the list: this topic has been discussed here many times and it does the list no good to start it all over again. Would be fine if you could respect private emails! John wrote: [...] Surely, however, the whole point of having a list is that not only can the original posters benefit from the wisdom of others but that others who have experienced similar problems can also benefit. Sending a private reply cuts them out. Does it not also mess up the list thread in the archives? It does now because you replied to the list. You should never answer emails that were sent to you as PM on a mailing list. If so, then your argument seems, on the face of it, to keep knowledge in restricted access rather than open mode. No, I think you didn't get the point. A list-reply functionality allows the user to decide whether the email should be public or private. And there are good reasons why some emails should be private. We could get rid of a lot of OT emails on this list! In summary: a) should the list be configured to support all the MUAs that do not have a list-reply functionality? This requires a reply-to pointing back to the list address to simplify usage. It will screw up the situation for anybody who has a MUA with list-reply functionality. b) should the list be configured to support MUAs with list-reply functionality? This means, the mailing list software is not allowed to set a reply-to header. It makes life easier for those who have a MUA with list-reply functionality but makes life a bit more complicated for all others. In addition, you can find some documents with technical arguments for not setting a reply-to header. The listmaster asked the mailing list members some time ago how to configure this list. The majority at that time decided to go with option b). So here we are. I can understand that some people are unhappy with the current situation but this is no reason to start the discussion about this topic again every week. I hope this clarifies the situation. Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [opensuse] Which Shell do I Have?
Anders Johansson wrote: On Tuesday 05 December 2006 21:15, Thomas Hertweck wrote: Hmmm... I would like to point out that this will (most likely) only print the default shell which is set in /etc/passwd (which should be fine in Kai's situation). If the current shell needs to be determined, it's better to use echo $0. ls -l /proc/$$/exe Could you please elaborate why you came up with this solution? I think this is not portable - for instance, our HP box here has no procfs, so your solution fails. An echo $0 should always work (as far as I know). Cheers, Th. -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]