Re: Intel (s775) Johannesburg DQ35JOE mATX
On 01/23/08 08:39:44PM +0100, Jonas Meurer wrote: On 23/01/2008 Steve Dobson wrote: You don't need a backport of the kernel packages. They are stand alone pieces of software that nothing depends upon and depend on nothing themselves (in the packaging sense; I know the system needs a kernel to run). You can just down load the kernel packages of your choice. Even though unstable kernel packages may work in etch, your statement is not true in general. Newer kernel needs newer udev at least, and I guess that some minimal glibc version is required as well. Actually the userland requirements have been pretty lax lately. According to the latest upstream kernel's git udev 081 is required and etch is already at 105. And just about any glibc should work because both the kernel and glibc have lots of compatibility to ensure interoperability. Hell there's not even a glibc version listed in the Changes file. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: rescue bootable cd ???
On 10/02/07 03:43:56PM -0400, Douglas A. Tutty wrote: On Tue, Oct 02, 2007 at 07:08:11AM -0500, helices wrote: Unfortunately, both systems on which I experienced such calamity ran lvm over software raid 5. In fact, both systems ran lilo, not grub; and everything was under lvm, including root and boot. Under these circumstances, there is specific configuration information missing, and that prevents the debian install cd, and knoppix, from being able to read my disks ; I thought that grub's support of LVM and any raid other than raid1 wasn't either supported or ready for prime-time? AFAIK there will never be any real RAID or LVM support in GRUB, it only sort of supports RAID1 because both volumes have the full filesystem on them. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: rescue bootable cd ???
On 09/28/07 05:26:27PM -0500, helices wrote: What do you do? The last few times I need to rescue something I used an Ubuntu disc that I had laying around, it doesn't include LVM but it uses a copy-on-write unionfs so that you can install packages onto the running LiveCD. Of course they're lost on reboot, but with a fast Internet connection that doesn't really matter. I would be surprised if Knoppix didn't already have similar functionality. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: rescue bootable cd ???
On 10/01/07 05:15:27PM -0500, C M Reinehr wrote: I've seen the term before but don't understand exactly what COW means, but I think I know what you're talking about. I had a failing disk drive last Friday and had to boot my system off of a Knoppix disk to see what I could salvage from my raid-1 array. Knoppix didn't seem to include the smartmontools package but I was able to use apt-get to install it on the fly and use it to diagnose which disk was failing. COW stands for copy on write, whenever a change is made to a block (or file maybe, I don't know what granularity unionfs uses) that block is copied somewhere else and the original left in place so that the update can succeed and the original still exists. Sort of like an LVM snapshot only instead of being backed by a LVM volume the changes are only stored in memory. But yea, the short answer is that it allows you to make changes and install software on read-only medium like CDs. =) Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Opinion question (Core2 Duo)
On 09/18/07 08:21:23AM -0300, Eduardo M KALINOWSKI wrote: Hamish Moffatt wrote: I only ever use my chroot to run mplayer occasionally now. (Especially with nspluginwrapper allowing Flash to be used from a 64-bit browser.) My two main desktops (home and work) are both 64-bit. Which leads me to the question: which video formats needs w32plugins yet? Since some time libavcodec can play wmv9 videos, and I'm not sure if I still need by 32bit mplayer or if its one less thing in the chroot (since openoffice.org is not necessary in a chroot anymore.) None that I know of, I haven't used 32-bit mplayer since I installed my 64-bit system. On a rare occasion some videos don't play well but I haven't taken the time to figure out if it's the video itself or some problem with the 64-bit build. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Kernel and Xen on an Intel Quad-Core Xeon E5320 processor
On 08/28/07 10:23:41AM +0800, GNUbie wrote: Hello all, I have an Intel Quad-Core Xeon E5320 processor running a Debian GNU/Linux Etch and I am confused on two things mainly: [1] Do I need to re-compile a new kernel to support SMP? Currently, it's running the stock linux-image-2.6.18-5-amd64 kernel. [2] I'm planning to install Xen on this machine so that I can create Xen images where the real network services will run on these images. Do I need to compile the newest version of Xen and related packages onto this machine from its project page or re-compile the Debian GNU/Linux Etch's Xen sources onto this machine or just use whatever is available from the Debian GNU/Linux Etch repository? Basically, I want that my Xen images will see that they're also running on a machine that comes with these four (4) processors and 4GB RAM. You can't give each VM (or domU in Xen terms) all of your memory so at the very least you'll have to rethink that part of your setup. As for the CPUs, Xen 3.0 and up does seem to support SMP domUs but I can't imagine it would be a very good idea to give multiple domUs all 4 CPUs. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64 X2 questions
On 08/25/07 01:22:50PM +0200, Michael wrote: Thanks Lennart, thanks Jim, for the good points. I think i can accept the matter of facts ;) It's generally a great fun to read your postings. Just let me say thx here for sharing your insight. That should apply to all those real freaks on this list. kr micha ps. I imagine quadcore (and beyond) may be able to solve what appears a little 'optical imbalance'. I would be interested to know if linux has specific schedulers for these archs but it's probably getting offtopic - i think quadcore is still rare at least in PC world yet. What is an 'optical imbalance'? No, Linux only has 1 CPU scheduler for all arches no matter how many CPUs are there. You can use cpusets to constrain processes to a specific set of memory and processors but they're never used automatically. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64 X2 questions
On 08/26/07 11:26:27PM +0200, Michael wrote: Just to avoid confusion...cpuset seems to be the kernel thing, and taskset is the related userspace tool ? Yes taskset is the userland tool for setting CPU affinity but it's only for setting CPU affinity while cpusets also restrict the nodes from which a process is allowed to allocate memory in addition to CPU affinity. What is an 'optical imbalance'? If one pane of the load-monitor is green and the other is black ;) It looks like it's not 'balanced' but you really clarified that issue now. Ah, ok. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Opinions on ext3 vs XFS vs reiserfs for LAMP server
On 08/23/07 04:07:46PM -0700, [EMAIL PROTECTED] wrote: Quoting Jim Crilly [EMAIL PROTECTED]: On 08/23/07 10:03:24AM -0700, [EMAIL PROTECTED] wrote: The problem of zeroing files of XFS still exists, however its not some mythical type of corruption. You'll only see it on files recently written to within seconds (say approx 60 secs) of a hard power off. If you can't risk it, or think you may have encounter the odd hard reset, ext3 might be a better choice. Actually it's been fixed as of 2.6.22: http://oss.sgi.com/projects/xfs/faq.html#nulls Of course that doesn't help you if you're using sticking with the kernel shipped with etch. I'm not so sure its fixed. I just tested with a sid samba box, running 2.6.22 kernel, and XFS filesystem. Connected to it via a WinXP box and copied a word doc file to it. Soft rebooted the samba box to make sure the file was sync'd to hard drive. Re-connected to samba share and opened the word document, added some text lines to it, saved and quit Word, then yanked the power out. Rebooted and re-connected to the samba share again only to find the file full of squares. Ext3 would have at least retained the original contents of the file. I tested the exact same thing again but waited 60 seconds after saving the file, and then yanked the power out. Upon a boot up, the file was intact and the save worked. So you still have about a 60 second window of newly written files and a power loss for data corruption, unless the program can sync it to disk before that. Well I'm only passing on what the XFS devs have said, all of my boxes are on UPSes so I rarely saw the issue anyway. But are you sure the squares you saw in the word doc were nulls? The FAQ page says that you can use the xfs_bmap command to see if it has any extents allocated and if it does then it would likely be another issue. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64 X2 questions
On 08/24/07 01:04:16AM +0200, Michael wrote: Lennart, probably don't have any good reason to. The scheduler does a great job Do you know if the kernel tries to balance load really equally ? For my 2core AMD64 it seems cpu0 has to be about 70% before cpu1 gets involved. Another idea is, i always see cpu0 is the most busy one. But wouldn't switching the 'most busy cpu' from 0 to 1 at certain intervals (say, 3 seconds) reduce heat ? It balances running processes but that's all it can do. So if you have 1 single-threaded process doing something it'll only run on 1 CPU and the other will sit idle. Trying to balance that one process between both CPUs would be horrendous for performance since it would constantly invalidate the cache on each CPU causing that data to be reloaded from main memory. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Opinions on ext3 vs XFS vs reiserfs for LAMP server
On 08/24/07 08:10:24AM -0700, [EMAIL PROTECTED] wrote: Quoting Jim Crilly [EMAIL PROTECTED]: I tested the exact same thing again but waited 60 seconds after saving the file, and then yanked the power out. Upon a boot up, the file was intact and the save worked. So you still have about a 60 second window of newly written files and a power loss for data corruption, unless the program can sync it to disk before that. Well I'm only passing on what the XFS devs have said, all of my boxes are on UPSes so I rarely saw the issue anyway. But are you sure the squares you saw in the word doc were nulls? The FAQ page says that you can use the xfs_bmap command to see if it has any extents allocated and if it does then it would likely be another issue. I'm not sure. I'm thinking the nulls thing is indeed fixed, but perhaps its still something different than the very nature of XFS. My understanding is limited of its mechanics, but since XFS will never journal data, only meta-data, then won't there always be the chance that a file can get corrupted on a timely power loss? AFAIK ext3 is the only one where data journaling is even an option and virtually no one uses it since it's off by default so XFS is no worse in that regard. Also since you were testing via Samba there's a chance that it's doing something funny that 'normal' Linux processes won't do when you save a file. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Opinions on ext3 vs XFS vs reiserfs for LAMP server
On 08/23/07 10:03:24AM -0700, [EMAIL PROTECTED] wrote: The problem of zeroing files of XFS still exists, however its not some mythical type of corruption. You'll only see it on files recently written to within seconds (say approx 60 secs) of a hard power off. If you can't risk it, or think you may have encounter the odd hard reset, ext3 might be a better choice. Actually it's been fixed as of 2.6.22: http://oss.sgi.com/projects/xfs/faq.html#nulls Of course that doesn't help you if you're using sticking with the kernel shipped with etch. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: resizing partition with Windows Vista
On 07/11/07 05:27:20PM -0500, Seb wrote: On Wed, 11 Jul 2007 18:04:10 -0400, [EMAIL PROTECTED] (Lennart Sorensen) wrote: [...] Some people use gparted from a livecd. Do NOT resize or move the windows boot partition however, unless you have a real vista CD around to boot to recovery mode and repair the boot files for vista. It will not boot if you change the boot partition in anyway (other than using the vista resize tool). That is the problem, the boot partition is the largest (137 out of the 160 Gb total), and can only be reduced to about 80 Gb!! The Vista system is optimized to make efficient use of the hard drive by creating 4 partitions with several unmovable parts, so that no matter how large your hard drive is, you can only reduce the partition used by Vista to about 3 x the space required. You never know how far M$ will go to make it easy for users to use their hardware... Those partitions were undoubtedly created by Toshiba and not Vista and as for the fact that Vista won't let you shrink the partition, by default it's pagefile is dynamically grown as necessary and with all of the other stuff that Vista does when running it makes sense for them to enforce a minimum amount of free space on the system partition. Depending on what Toshiba put on the other 2 partitions (assuming one is Vista and one is recovery) you should be able to delete them and have Vista run just fine. But if it were my laptop I would use whatever tool Toshiba provides to make a recovery disk, delete all the partitions and start over with just Debian and then run Windows inside VMWare or whatever virtualization you prefer. Although you probably won't be able to use the restore disk in the virtualized system so you might have to jump through some hoops to get a working Vista system that way. -- Seb Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 32-bit vs AMD64 on Opteron for LAMP server
On 07/07/07 07:36:26PM -0400, Roberto C. Sánchez wrote: On Sat, Jul 07, 2007 at 05:39:54PM -0400, Jim Crilly wrote: On 07/07/07 04:45:57PM -0400, Roberto C. Sánchez wrote: The stock Debian kernels are configured like this: CONFIG_HIGHMEM4G=y # CONFIG_HIGHMEM64G is not set CONFIG_HIGHMEM=y So, if you have a machine with 4-64 GB RAM, then a custom kernel is in order. Of course, as far as the BIOS goes, if the machine supports more than 4 GB RAM, then the BIOS should as well. After all, why would someone manufacture a machine that can handle more than 4 GB RAM and then put in a BIOS that cannot? No, even with just 4G you need CONFIG_HIGHMEM64G because to access the memory from ~3.5G-4G you need to remap it above the 4G mark since those addresses were stolen by the various hardware components in your system so you need a kernel able to address 4G. Please read again my first sentence. You and I are in agreement on this, just saying it in different ways. Whoops, yea I did misread that. But a custom kernel shouldn't be necessary because on 32-bit systems the 686-bigmem kernel has CONFIG_HIGHMEM64G and on 64-bit systems there's nothing special needed. Regards, -Roberto Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 32-bit vs AMD64 on Opteron for LAMP server
On 07/08/07 09:49:32AM -0500, Neil Gunton wrote: Jim Crilly wrote: For 32-bit systems only if the kernel is compiled with CONFIG_HIGHMEM64G enabled so you need one of the bigmem kernels. And the BIOS on the machine has to support remapping the lost memory above the 4G mark, if it won't do that for you there's nothing you can do to get access to that memory. The per-process limit will be 3G since 4G is the max addressable and 1G of that space is reserved for the kernel. And part of the 3G will be used for the binary itself, shared libraries, mmap'd files, etc so you'll never even get the full 3G out of a single process. Ok, here's another thought: It's easy to get into a purist frame of mind with this where you end up obsessing over numbers like 3 GB or 4 GB or 64 GB. But taking a step back, does any real world process, especially in the LAMP stack, actually need that amount of memory? I know MySQL can probably use whatever you want to throw at it, for index buffer caching particularly, but I can only envisage giving MySQL 3 GB if the total system RAM was a lot more than that - probably at least 6 GB, and I don't plan on doing that soon with this server. Well I can't really comment on how useful it would be because that would be dependent on the usage patterns of the installation and I've never had to deal with a LAMP installation of any real size. But the memory limits are virtual, not physical. So if you had planned on giving MySQL 3G for anything you would need a larger VM space because a chunk of it will already be used. I'm not sure if this was mentioned but another option would be to install the 32-bit i386 distribution but run a 64-bit kernel, that way each 32-bit process would have 4G of VM since the kernel wouldn't have to share their address space and you would also have the option of running some select 64-bit binaries if you find that something needs more VM. I didn't even know you could do that! Sounds interesting, though. Seems a little weird mixing the two modes, because I would have thought that processes talking to the kernel would need to be using the same underlying libs and agree on what size an integer is, for example. But evidently it's possible, so thanks for the idea. It's essentially the same concept as installing the 64-bit release and then putting a 32-bit chroot on that box to run things like flash, the same kernel is executing both sets of executables. Most processes talk to the kernel very little and do so in via well defined syscalls so there's no problem running 32-bit binaries on a 64-bit kernel. Occasionally something that does do odd things with the kernel, like the iptables userland stuff, can have issues but those are special cases. The kernel doesn't use any libraries itself so that's not a problem, but yes you do need all the right versions and builds of any dependencies. So if you have a 32-bit userland and decide that you want to run a 64-bit MySQL you need to build 64-bit versions of MySQL and all of the libraries that it needs before it'll run. I'm not sure if you could setup a 64-bit chroot on the 32-bit system to make that simpler but if so that would take away the pain of having to build anything. /Neil Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 32-bit vs AMD64 on Opteron for LAMP server
On 07/06/07 10:40:11PM -0500, Neil Gunton wrote: Adam Stiles wrote: You won't be able to use all of your 4GB RAM with a 32-bit kernel. A 32-bit processor only has 4GB of addressing space, and that has to be shared between memory and peripherals. Really? I thought that the only limitation was on individual processes not having more than 4GB available, but the entire system as a whole could address a lot more than that. But I could be wrong. For 32-bit systems only if the kernel is compiled with CONFIG_HIGHMEM64G enabled so you need one of the bigmem kernels. And the BIOS on the machine has to support remapping the lost memory above the 4G mark, if it won't do that for you there's nothing you can do to get access to that memory. The per-process limit will be 3G since 4G is the max addressable and 1G of that space is reserved for the kernel. And part of the 3G will be used for the binary itself, shared libraries, mmap'd files, etc so you'll never even get the full 3G out of a single process. I'm not sure if this was mentioned but another option would be to install the 32-bit i386 distribution but run a 64-bit kernel, that way each 32-bit process would have 4G of VM since the kernel wouldn't have to share their address space and you would also have the option of running some select 64-bit binaries if you find that something needs more VM. The only bad thing about that is you would have to compile whatever 64-bit stuff you need on your own since the only 64-bit packages in i386 seem to be the kernel and a handful of libraries. Thanks again, /Neil Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 32-bit vs AMD64 on Opteron for LAMP server
On 07/07/07 04:45:57PM -0400, Roberto C. Sánchez wrote: On Sat, Jul 07, 2007 at 04:40:00PM -0400, Jim Crilly wrote: For 32-bit systems only if the kernel is compiled with CONFIG_HIGHMEM64G enabled so you need one of the bigmem kernels. And the BIOS on the machine has to support remapping the lost memory above the 4G mark, if it won't do that for you there's nothing you can do to get access to that memory. The stock Debian kernels are configured like this: CONFIG_HIGHMEM4G=y # CONFIG_HIGHMEM64G is not set CONFIG_HIGHMEM=y So, if you have a machine with 4-64 GB RAM, then a custom kernel is in order. Of course, as far as the BIOS goes, if the machine supports more than 4 GB RAM, then the BIOS should as well. After all, why would someone manufacture a machine that can handle more than 4 GB RAM and then put in a BIOS that cannot? No, even with just 4G you need CONFIG_HIGHMEM64G because to access the memory from ~3.5G-4G you need to remap it above the 4G mark since those addresses were stolen by the various hardware components in your system so you need a kernel able to address 4G. Regards, -Roberto Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64 show stoppers
On 01/16/07 07:42:29PM -0600, Karl Schmidt wrote: Add your show stopper to this thread... First on the list is the java browser pluging - the black down java sort-of-works(tm) (not a Debian package) Since Sun's GPL'ing Java that'll probably be taken care of in the future, but the chances of it making it into Etch are slim. But frankly I don't see the problem. Next is flash - you can sort of workaround it by using switftfox (comes with bugs)(not a Debian package) Same here, I like the fact that my browser doesn't have flash. I do keep a 32-bit chroot around 'just in case' I need it, but that's pretty rare. But the biggest problem for me is that there isn't a AMD64 nvu package. The author abandoned it to work on a replacement package, but in the mean time there isn't a good WYSIWYG html editor for web development. (Sorry, but quanta just isn't in the same league and OO makes bloat-code out of clean html). nvu doesn't exist at all in Debian as far as I can tell, so that's not AMD64-specific. But there does seem to be nvu packages in Ubuntu and they even have AMD64 builds, so you could try building it yourself or rebuilding the Ubuntu source packages if you really want it. IMHO etch as a amd64 server is fine - but the Desktop has holes The Linux desktop in general has holes, but IMO Etch itself is fine. I've been using the AMD64 port on this machine since I built it a few months ago and haven't had a single problem. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64 show stoppers
On 01/17/07 01:47:00PM +1100, Alex Samad wrote: On Tue, Jan 16, 2007 at 09:41:38PM -0500, Jim Crilly wrote: On 01/16/07 07:42:29PM -0600, Karl Schmidt wrote: Add your show stopper to this thread... First on the list is the java browser pluging - the black down java sort-of-works(tm) (not a Debian package) Since Sun's GPL'ing Java that'll probably be taken care of in the future, but the chances of it making it into Etch are slim. But frankly I don't see the problem. Majour problem with so much java being used in web sites, I find it really annoying - the solutions I use is the blackdown - there is a apt repository and the gcj plugin - still early days I can't remember the last site I saw that required Java. Actually I can, but it was an internal site at the last company I worked for and it didn't work in anything but IE anyway. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: i386 or amd64?
On 01/02/07 12:02:37PM -0600, Mike Reinehr wrote: Greg, On Monday 01 January 2007 18:25, Greg Madden wrote: On Mon, 09 Oct 2006 21:30:08 +0200 Thomas Steffen [EMAIL PROTECTED] wrote: On 10/8/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: Other than non-free stuff like flash and openoffice, is there anything in the main section of i386 that isn't in the main section of AMD64? A few minor things are missing, such as memtest86, partimage, and of course wine. Also note that most java-packages come without the plugin in 64bit. In other words, what are people having to use a chroot i386 for? Closed software such as acrobat reader, flash, skype, vmware, picasa, googleearth etc. Most can be installed without a change root, but that is a bit of a struggle. Since you have only 1 GB of RAM, you can go with i386 without any obvious penalty. With amd64 you can get a little bit more performance, but you are also more likely to run into problems, such as the ones mentioned here. Thomas FYI, the latest VMware runs on 64 bit linux hosts. Which VMware? Just this past Saturday I downloaded tried to install Workstation 5.5.5-29772 and the install failed with a list of missing libraries (32-bit, I presume, although I haven't had time to investigate). I don't know about VMWare Workstation, but I know Server works although the UI is 32-bit so you need at least ia32-libs installed. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel upgrade problem
On 12/31/06 06:40:52PM -0600, Russ Cook wrote: I had already checked with 'df' before posting, but I attach the results in case it offers a clue. Thanks for the reply, Russ Script started on Sun 31 Dec 2006 06:38:39 PM CST Filesystem 1K-blocks Used Available Use% Mounted on /dev/hda1 255912200418 41841 83% / tmpfs 1674944 0 1674944 0% /lib/init/rw udev 1024064 10176 1% /dev tmpfs 1674944 0 1674944 0% /dev/shm /dev/hda9181745360 90136804 82376404 53% /home /dev/hda8 369000 8355340985 3% /tmp /dev/hda5 4807056 2657124 1905748 59% /usr /dev/hda6 2885780 1257440 1481752 46% /var Script done on Sun 31 Dec 2006 06:38:39 PM CST Did you check the inode count? If not, see what 'df -i' says. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Kernel Configuration Question
On 12/22/06 12:10:41PM -0600, Mike Reinehr wrote: I hate to answer my own posting but it belatedly has occurred to me that perhaps it's not possible to mount a root partition using LVM without an initrd.img. I've booted without an initrd.img before I've used LVM before, but not with the root partition as part of the logical volumes. Yes, no, maybe? That's pretty much it, you need to run the LVM tools (vgchange I think) to scan for and setup the logical volumes. There is no code in the kernel to do that for you so you have to use an initramfs image if your root is on LVM. But why go through all of that trouble to not use one? The only burden it puts on you is to run 'update-initramfs -u -k kernel-version' on of the off chance that you change something that also needs to go in the image, normal updates to things like LVM tools, udev, etc should update it for you. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Kernel Configuration Question
On 12/23/06 11:25:25AM -0600, Mike Reinehr wrote: On Saturday 23 December 2006 09:35, Jim Crilly wrote: On 12/22/06 12:10:41PM -0600, Mike Reinehr wrote: I hate to answer my own posting but it belatedly has occurred to me that perhaps it's not possible to mount a root partition using LVM without an initrd.img. I've booted without an initrd.img before I've used LVM before, but not with the root partition as part of the logical volumes. Yes, no, maybe? That's pretty much it, you need to run the LVM tools (vgchange I think) to scan for and setup the logical volumes. There is no code in the kernel to do that for you so you have to use an initramfs image if your root is on LVM. But why go through all of that trouble to not use one? The only burden it puts on you is to run 'update-initramfs -u -k kernel-version' on of the off chance that you change something that also needs to go in the image, normal updates to things like LVM tools, udev, etc should update it for you. Jim. Thanks for confirming this. I think I may have read something about this last year when I first researched LVM but then forgot. As for not using an initrd.img, long ago I became a confirmed follower of the KISS theory of operations (Keep It Simple, Stupid) and was just trying to pare my kernel of any unnecessary pieces. But, as you say, it's not that much trouble to maintain an initrd.img. Exactly, it's virtually 0 maintenance unless you're doing really odd, complex things in your initramfs and even then once you set it up and put the files under /etc/initramfs-tools/ it'll keep working. A decent example is this, I setup this box with some dm-crypt block devices and by default the generic aes module is used, to switch to aes_x86_64 all I had to do was put the module name in /etc/modules, update my initramfs and reboot, if they had been compiled in statically I would have been stuck recompiling my kernel for that. And with the kernel people wanting to push more and more device discovery and setup to userland it's going to be unavoidable at some point anyway. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 2D,3D,nvidia,nv?
On 12/17/06 09:31:18PM +, Paul Brook wrote: On Sunday 17 December 2006 21:03, Brian R. Whitecotton wrote: IMHO I don't see the point in having a GeForce 7300 GT unless you are at least enabling its power/capabilities. A 7300GT is a fairly bottom-of-the-line card. It's the cheapest card I've seen that has dual-link DVI connectors (required for big, high resolution monitors). The 2D nv driver is fine but the nvidia driver is better. Your definition of better is very different to mine. The open source does everything I need (high resolution, fast 2d, video). The binary driver doesn't work at all under Xen, and locks up periodically on half my machines. The binary driver works fine under Xen on i386 and people have gotten it to work with the AMD64 Xen kernels on the nvnews.net forums so it is possible. And the last time I tried the OSS nv driver it didn't do Xv at high resolutions[1], the image quality was noticably lower with a 24 bit desktop and it was a lot slower even in the 2D arena; for instance switching desktops would take a second or two with the nv driver but with nvidia it's almost instantaneous. Obviously both drivers will work better or worse on different hardware so neither is a clear winner in all cases and everyone needs to decide on their own which to use. I'm not advocating the use of the closed driver at all, hell it's caused me number of problems on my notebook but on my desktops it's been nearly flawless. And I would be ecstatic if nv or the nouveau driver would get to the point where just their 2D is as good as the binary driver, but right now they lag behind pretty badly IMO. Jim. [1] https://bugs.freedesktop.org/show_bug.cgi?id=474 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: etch RC1 installer
On 12/13/06 03:23:41PM +, Paul Brook wrote: The nvidia binary driver does DRI openGL. The free driver is 2D only. I also believe the free driver doesn't do video playback acceleration while the binary one does. The open source driver does support video overlays, and seems to provide hardware acceleration (I see a 10x reduction in CPU usage when xv is enabled). What about at high resolutions though? According to this[1] bug report at fd.o it's still broken and won't be fixed until at least Xorg 7.3. And the last time I used the nv driver it was noticably slower than the binary nvidia driver even at normal 2D desktop stuff, not that it was unusable or anything but the speed difference is significant. Jim. [1] https://bugs.freedesktop.org/show_bug.cgi?id=6151 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: etch RC1 installer
On 12/13/06 10:18:20PM +, Paul Brook wrote: What about at high resolutions though? According to this[1] bug report at fd.o it's still broken and won't be fixed until at least Xorg 7.3. And the last time I used the nv driver it was noticably slower than the binary nvidia driver even at normal 2D desktop stuff, not that it was unusable or anything but the speed difference is significant. I'm running amd64 Debian unstable at 1600x1200 on a 128Mb GF4200ti. Up until a couple of months or so ago it was painfully slow. I couldn't play video above about 640x480, and other things (firefox, kpdf) were noticeably slower than they should be. Interesting, when I ran into the problem it was with the i386 port so that may make a difference. I had assumed that the AMD64 port would have the same limitations since it's a video thing but maybe that's not true. If I get a chance I'll try switching this machine over to the nv driver and see if Xv output in mplayer works, although it's a completely different machine so it won't be a very good test but I'd still be interested in knowing if it works or not. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Modulized or monolithic kernel on notebooks ?
On 11/30/06 10:06:44AM +0100, Hans-J. Ullrich wrote: Hi all, I know, there was a lot of discussion about kernel-builds. And I really do not want start flamewars. So here are my little questions: What do you think (with the the focus on speed): 1. Does it make sense to compile a kernel with all modules built in, for the hardware which is always present on the target notebook (maybe desktop-px, too) ? I.e. sound, controller, filesystem, pcmcia-port, usb, sd-card-reader, wlan etc. etc, everything, which cannot be exchanged. No, especially considering that USB is one of the worst offenders in the stops my notebook from going to sleep group and the best way to work around that is to unload the USB modules. The only benefit you'll get is that you won't need an initramfs image to boot and the value of that's debatable. And if you ever want to use uswsusp you'll need an initramfs image anyway. 2. Does this improve speed especially on 64-bit-systems ? If there is any difference I guarantee it'll be so small to be well within the margin of error of any benchmarks. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Modulized or monolithic kernel on notebooks ?
On 11/30/06 10:07:43AM +, A J Stiles wrote: I know of some people who like to install monolithic kernels (and disable module loading) on servers; but that's done for security reasons, not for speed. And those gains are dubious at best, kernel memory can still be altered via /dev/kmem or /dev/mem on most distributions, I think the only one to include any patches to mitigate that is Fedora but I'm not sure. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel without smp
On 11/22/06 02:37:19AM +0100, sigi wrote: On Tue, Nov 07, 2006 at 03:23:36PM -0500, Jim Crilly wrote: Have you tried booting with the kernel parameters 'nosmp' or 'maxcpus=1'? Just for your information: this two parameters do _not_ work - they both seem to have no affect. ifup freezes the keyboard everytime I try to connect my wlan-card. But the badest thing is: after my last apt-upgrade no available rt2570 wlan-card-module can connect to the internet! A fiew days I could use a daily build from the developers-website [1], but since my last upgrade this doesn't work anymore. For now it's impossible for me to connect to the internet with my amd64-machine - and that because there are only smp-enabled linux-images out there for amd64 very very bad! No affect in that both CPUs are always detected or just that the box still locks up? Although either way your only real option is to compile a custom kernel for UP. Sounds like the rt2xxx developers need to get their act together and fix their locking problems, things will only get worse as more and more dual-core notebooks are released. I've actually had similar issues with the rt2500 driver on my notebook since it has HT, but since I was already using a custom kernel to apply some 3rd party patches I just disabled SMP to work around it. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel without smp
On 11/22/06 03:28:25AM +0100, sigi wrote: On Tue, Nov 07, 2006 at 03:23:36PM -0500, Jim Crilly wrote: No affect in that both CPUs are always detected or just that the box still locks up? If I remember right, the nosmp-option produced a kernel-panic during boot-time, and maxcpus=1 froze the keyboard while trying to get my IP with ifup. Although either way your only real option is to compile a custom kernel for UP. I tried this yesterday, but dpkg didn't install my custom kernel... I don't know, what the problem was - I tried the 'official' debian way, and only disabled smp-support from the debian-kernelimage config-file. But dpkg could not install it You did use make-kpkg to build it first, right? Sounds like the rt2xxx developers need to get their act together and fix their locking problems, things will only get worse as more and more dual-core notebooks are released. Possibly I should write a new post into their forum, that they think about this in the future?!? AFAIK they've known about it for a long time, it's been an issue for as long as I've had my rt2500 card. But it can't hurt to remind them. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Help: NO longer boots ;
On 11/11/06 08:29:02PM -0600, helices wrote: I have a dual Opteron box (Odin) that went down while I was gone, due to an extended power failure. When it comes up, it fails to complete boot. Please, help me get this system running again. uname -a Linux (none) 2.6.17-2-amd64 #1 SMP ... Bootup sequence from obvious problems: ... Success: loaded module raid10. Done. device-mapper: 4.6.0-ioctl (2006-02-17) initialised: ... 1 logical volume(s) in volume group VG2 now active 7 logical volume(s) in volume group VG1 now active Done. stdin: error 0 Begin: Running /scripts/local-premount ... kinit: name_to_dev_t(/dev/mapper/VG1-swap) = dm-3(253,3) kinit: trying to resume from /dev/mapper/VG1-swap Attempting manual resume kinit: No resume image, doing normal boot ... Done. Usage: modprobe ... ... Begin: running /scripts/init-bottom ... mount: Mounting /root/dev on /dev/.static/dev failed: No such file or directory Done. mount: Mounting /sys on /root/sys failed: ... mount: Mounting /proc on /root/proc failed: ... Target filesystem doesn't have /sbin/init ... (initramfs)_ blinking cursor First see if anything is in /root, it looks like it's attempts to mount the root filesystem there so if something is there it'll tell you if it just mounted the wrong filesystem or if there's a real problem with your root filesystem. If there's nothing at all there just run 'mount' and see if it thinks something is mounted there, if not look in /dev/mapper and see if there's a LVM volume for your root, considering that your swap is called VG1-swap it'll probably be VG1-root or VG2-root. If it's there try mounting it, although I'm not totally sure all of what capabilities you'll have in the initramfs. It might also be a good idea to look around for a Knoppix, Ubuntu LiveCD, etc so that you can have more tools handy in case there is a real problem with your disks. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel without smp
On 11/07/06 03:07:47PM +0100, sigi wrote: Hi, as I noticed today, for my wlan-card-module, I need a kernel without smp-support enabled... This was the reason, my new rt2570-module always locked my keyboard: the prebuilt debian kernel I'm using (2.6.17-2-amd64) has smp enabled. Now, to resolve the problem, do I have to build my own kernel (what I not really want), or is it enough to install the package linux-image-2.6-amd64-k8? I didn't find any kernel *amd64-k8 for etch, which existed earlier, so I think that linux-image-2.6-amd64-k8 (as transitional package) will only update to the latest generic amd-kernel available? So, only chance to build my own one to use my rt2570-module? Have you tried booting with the kernel parameters 'nosmp' or 'maxcpus=1'? Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64-generic doesn't see all 4GB RAM?
On 11/07/06 11:07:56PM +0100, Daniel Tryba wrote: Well, I personally have been trying debian kernels for a couple of weeks. A machine that has been running Debian/unstable with custom kernels for the last couple of years and always has been very stable, now is not detecting the soundcard every couple of reboots. And since I can't (mainly to lazy to figure out how to) get VMWare Server to work anymore (I can't seem to find the correct headers for the installed kernel) I'm making the same conclusion I made a couple of years ago: custom kernels are much more reliable and easier. I've been using the pre-built kernels since I got this AMD64 box without any hardware detection issues. If your soundcard disappears every couple of reboots I would be more suspicious of the hardware than the kernel. As for VMWare, the Ubuntu wiki listed the prereqs that I needed and after they're installed you can install VMWare just like normal. https://help.ubuntu.com/community/VMware_Guide%3a_Installing_VMware_Server_on_Ubuntu_6%2e06_LTS_amd64 Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: AMD64-generic doesn't see all 4GB RAM?
On 11/08/06 12:02:36AM +0100, Daniel Tryba wrote: On Tue, Nov 07, 2006 at 05:38:50PM -0500, Jim Crilly wrote: I've been using the pre-built kernels since I got this AMD64 box without any hardware detection issues. If your soundcard disappears every couple of reboots I would be more suspicious of the hardware than the kernel. This is contradicted by the custom kernels not having this problem. True, so if you're using the same version of the kernel there's either a patch upstream fixing something that's not in the Debian kernel or it could be an odd timing issue if you're compiling the sound driver in statically but that seems unlikely. As for VMWare, the Ubuntu wiki listed the prereqs that I needed and after they're installed you can install VMWare just like normal. https://help.ubuntu.com/community/VMware_Guide%3a_Installing_VMware_Server_on_Ubuntu_6%2e06_LTS_amd64 Thanks for the link, but: -the machine having problems is not running a 64bit kernel -I already installed kernel headers and tried pointing the vmware-config to many places without success. Again these troubles aren't present when running custom kernels. 2 out of the 3 machines I'm tried to run deb. kernels on failed. The third on is running just fine. Well I don't know what to say about that, I've installed VMWare on numerous machines with custom and Debian kernels and the only time I've run into problems with the VMWare kernel modules was with custom kernels mostly because of some 3rd party patches. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: motd
On 10/30/06 07:04:08PM -0500, Douglas Tutty wrote: On Mon, Oct 30, 2006 at 07:55:07PM +0100, Manolo D?az wrote: Mike Reinehr wrote: On Monday 30 October 2006 11:59, Douglas Tutty wrote: I'm running Etch on my new box and getting it set up. /etc/motd now points to /var/run/motd and there's a file /etc/motd.tail Hi, From the getty manpage, section ISSUE ESCAPES. s: the system name, the name of the operating system. n: the nodename of the machine, also known as the hostname. r: the release number of the OS, eg. 1.1.9. v: the version of the OS, eg. the build-date etc. m: the architecture identifier of the machine, eg. i486 Regards, Manolo. How does /etc/issue relate to /etc/motd? /etc/issue and /etc/issue.net are printed before the login prompt, /etc/motd is printed after login. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian AMD64 boots only at random: how to use labels/fstab/grub
On 10/02/06 09:40:00AM -0700, Bob McGowan wrote: Sorry, I forgot that bit ;( The command 'mkswap -L label device' will create a swap area and add a label for it. This should probably be done from a rescue disk environment, so you don't confuse the running kernel with changes to its swap area. I'm not aware of a separate tool to label a swap area without reformatting it, though such may well exist. Just do 'swapoff /dev/whatever' first and you'll be fine. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Problem with recent udev upgrade (0.097-1)
On 08/19/06 12:55:57PM +0200, Sylvain Archenault wrote: Hi, This morning, I upgraded udev to 0.097-1 from 0.093-1. After reboot, I've got problem with my usb mouse and my usb wifi adapter. I solved those problems by adding the corresponding module to /etc/modules. I still have problems with my sound card (Realtek ALC850 on Asus K8N-E Deluxe board), I can't find the correct module. I was thinking it was snd-intel_8x0, but it doesn't works fine. On boot time, i noticed a lot of errors coming from udev, telling an error occurred while look up for group audio (but also nogroup), the reason is illegal seek (I don't find those lines in log files). Those errors might the cause of my problem. I found nothing interesting on the internet, neither in udev bugs. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=383555 I aslo tried to downgrade to 0.093-1, but I still got the same errors at startup. As someone else mentioned it's a bug in a klibc function that udev just started using, if you downgrade udev it won't fix anything unless you regenerate your initramfs image. Or you could just wait a day or so as an updated klibc has been uploaded today. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ndiswrapper screwed up again?
On 08/09/06 06:24:31PM +0200, Erik Mouw wrote: On Wed, Aug 09, 2006 at 12:55:11PM +0200, Hans wrote: Am Mittwoch, 9. August 2006 13:56 schrieb Rafael Rodríguez: I'm using 2.6.17-ck1, and the driver still doesn't work for me :) I haven't really thought about it that much due to its alpha state, but it loads and detects my wlan BUT won't associate with any AP. That is weired. Just try to add the gateway of the AP in /etc/network/interfaces. This worked for me. And please check, if the correct ethx is wlan, if you have wo network-devices. Remember: The Broadcom card is seen as ethX not as wlanX ! You can change that with ifrename. Although using ifrename is supposedly racy now and should be avoided if possible, if you want to rename devices you should use udev rules like those automatically generated in /etc/udev/rules.d/z25_persistent-net.rules Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: reiserfs/md1/failure/threads
On 07/18/06 05:27:17PM +0200, Erik Mouw wrote: On Tue, Jul 18, 2006 at 04:15:42PM +0200, Goswin von Brederlow wrote: Francesco Pietra [EMAIL PROTECTED] writes: The most recent FS is generaly the one with the most unfound bugs left and often a lot of design kinks that remain to be fixed. Something like ext2 on the other hand has all the bugs and kinks worked out over the years and there is very little new code that could go wrong. Ext3 should do as well. The same team that maintains ext2 also maintains ext3. Ext3 is like ext2, but with journaling and directory indexing. Bug fixes from ext3 get backported to ext2. Ideally yes, but a lot of times bug fixes and other code changes don't get backported. From what I've seen very few people think Hmm, I should check ext2 for that too when they make a change to ext3. This isn't a knock on ext3, it's been extremely reliable in the places that I've used it and I would definitely recommend it over reiserfs any day, I'm just saying that ext2 and ext3 aren't really the same any more. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: XFS, EXT3 or some other?
On 06/17/06 07:39:16PM +0200, Thomas Steffen wrote: On 6/16/06, Hemlock [EMAIL PROTECTED] wrote: I've read some articles googling for xfs, ext3 and jfs and such. Leaning towards xfs maybe? Don't forget reiserfs. In my experience it works very well. It is supposed to be more space efficient. On the other hand it also seems to have performance problems in a few cases, that the other file systems don't have. Generally it is quite fast, though. Every time I've tried reiserfs it's resulted in problems, not usually right a way, but eventually. The last time the corruption wasn't even detected by the kernel driver or reiserfsck, both said the fs was fine but any time I accessed a certain file the screen would blank and the box would hang, I had to hookup a serial cable to figure out that it was even reiserfs. Ext3 is certainly a safe choice, and with the directory hash it should give really decent performance. One big advantage is that you have so many ways to access it (rescue disk, Windows driver etc). XFS is very fast in my experience, but it did have some issues on AMD64. There where a number of recent kernel patches, e.g. log recovery is now compatible between 32bit and 64bit. I also found that it has a very annoying tendency of leaving corrupted files around after a crash (which I never had with ext2, ext3 or reiserfs). Grub did not support XFS, although that might be fixed now. There was also talk about problems between NFS and XFS, but I didn't not follow that. AFAIK grub will never work with /boot on XFS because of where the XFS superblock is, it's not too big of a deal to make a small ext2 /boot though. I'm using XFS on i386, sparc64 and Alpha without any issues, but I don't have any AMD64 system to put it on yet. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/23/06 02:34:35PM +0200, H. Wilmer wrote: Jim Crilly wrote: More stuff is configurable without a recompile. If I add/change hardware I don't have to do anything unless it's something required for booting and even then updating my initrd is simple. And I have run into cases where reloading modules will fix things, Ok, that are advantages to be made use of. But when hardware is not about to be changed and it's for a server that should just run the way it is reliably, what's the advantage of using modules for things required to boot anyway? MOTT, you could not even unload those modules. Even with servers there's a good chance you'll have to replace hardware and with the way companies tend to change chipsets and revisions without changing names it's still safer to just use modules. IMO You don't gain anything by compiling them in statically. I'm looking at it from the other direction, with initrd so easy to create, what's the advantage of not using modules? Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/23/06 04:42:26PM +0200, H. Wilmer wrote: Jim Crilly wrote: anything by compiling them in statically. I'm looking at it from the other direction, with initrd so easy to create, what's the advantage of not using modules? Ok, maybe it's just me thinking that I gain reliability whith some things compiled in and others not, in an attempt to take advantage of both approaches :) GH Not in any way that I can think of, the only thing it does is simplify booting a very little bit by not requiring an initrd. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/22/06 01:51:37PM +0200, H. Wilmer wrote: Jim Crilly wrote: There's no such module; I think SCSI support is compiled into the kernel. That would probably be a valid assumption. Yeah, I'd say I did that :) Well to even attempt reloading anything you need it as a module, infact I prefer to make everything possible a module and use an initrd. But in this case it wouldn't make a difference since you can't umount the filesystem to be able to reload the module anyway. True --- and then, what's the advantage of using so many modules? If something is compiled into the kernel, there cannot be any trouble with an inability to load it. So I prefer to compile in everything that is required anyway to get things running and cannot be removed. GH More stuff is configurable without a recompile. If I add/change hardware I don't have to do anything unless it's something required for booting and even then updating my initrd is simple. And I have run into cases where reloading modules will fix things, of course it's not possible for a few things like the filesystem and store driver of the root fs, but most other things can be reloaded if you kill the processes using them. I've had a number of times where reloading a NIC driver reset the NIC and fixed a problem and I've had to switch between ALSA and OSS sound drivers more than I'd like to admit. I don't see the benefit of compiling things in statically, you're limiting yourself too much that way IMO. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/19/06 01:45:34PM +0200, H. Wilmer wrote: Jim Crilly wrote: Reloading a SCSI host driver (i.e. gdth, aic7xxx, etc) will increase the counter, reloading the SCSI core itself (scsi_mod) will reset the counter back to zero. There's no such module; I think SCSI support is compiled into the kernel. That would probably be a valid assumption. I would also have compiled the gdth support in, but unfortunately it was only possible to compile it as a module. That in turn required to use an initrd.image :( Well to even attempt reloading anything you need it as a module, infact I prefer to make everything possible a module and use an initrd. But in this case it wouldn't make a difference since you can't umount the filesystem to be able to reload the module anyway. Are you sure you're using gdth for your root? I thought it was just for SCSI controllers, I didn't know any SATA controllers used chipsets supported by that driver. Yes, the Vortex is an SATA RAID controller and needs the gdth module: :02:02.0 SCSI storage controller: Adaptec ASC-29320A U320 (rev 10) :03:01.0 RAID bus controller: ICP Vortex Computersysteme GmbH GDT NEWRX There are no other disks than those attached to the RAID controller in the server. The only device on the Adaptec is a tape changer. I can recommend the Vortex controller, no problems with it in about two years. It's running a RAID5, 1TB on 4 disks plus one spare. If a disk fails, it will beep, so you _will_ notice. You replace the disk and that's all. --- I've had a disk failing in another server that also has a Vortex after about 3 months. It beeped and kicked in the spare; the replaced disk is now the spare, so it gets away with only one rebuild. 3wares are a little cheaper and work also (no problems except for a disk starting to fail in about 3 years, running RAID1 with two IDE disks), but they don't have a beeper. A beeper is definitely worthwhile. GH I was under the false impression that the gdth driver was just for some older legacy cards, the card definitely sounds nice I'll have to consider one next time I need a RAID controller. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/18/06 02:41:31PM +0200, H. Wilmer wrote: Jim Crilly wrote: It seems to be normal, but I'd probably still say it's a bug in the SCSI system. It is possible to reset the number by reloading the scsi_mod module, but you have to umount all of the SCSI filesystems to do that so it's not a great solution. Unloading the SCSI module is what I did, and the counter is increased each time I reload the module. But there's also the gdth module for the SATA RAID controller, generating SCSI devices, which cannot be unloaded unless the server could run diskless ... GH Reloading a SCSI host driver (i.e. gdth, aic7xxx, etc) will increase the counter, reloading the SCSI core itself (scsi_mod) will reset the counter back to zero. Are you sure you're using gdth for your root? I thought it was just for SCSI controllers, I didn't know any SATA controllers used chipsets supported by that driver. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: counting scsi hosts
On 05/17/06 01:04:59PM +0200, H. Wilmer wrote: Goswin von Brederlow wrote: when unloading and reloading SCSI modules, the number of SCSI hosts is increased like this: Also happens on usb every time you unplug and replug a harddisk. Very anoying. Thanks! At last I'm still on the save side since it's normal. GH It seems to be normal, but I'd probably still say it's a bug in the SCSI system. It is possible to reset the number by reloading the scsi_mod module, but you have to umount all of the SCSI filesystems to do that so it's not a great solution. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: new kernel too big for lilo
On 12/31/05 12:37:12AM +0900, Craig Hagerman wrote: Thanks for all the feedback. I tried installing a new kernel the 'debian way': % make menuconfig % make-kpkg followed by: % dpkg -i kernel_name.deb which did everything automatically. Then I restarted ... to find I had no GUI and no internet. I realized that the automatic install had renamed my old kernel by appending .old to the name. I was able to add THAT to lilo.conf by hand and successfully reboot. I would have thought that both would be available in lilo after the update. But I don't use LILO and I don't let make-kpkg touch my bootloader so I could be wrong. I am sure most of you will disagree with me but this is one area where I do NOT like doing things the debian way. Compiling and installing a kernel isn't something I do everyday but it is something that can mess up a system. I don't know what is automagically being done behind the scenes and I am very uncomfortable with that. I would much rather follow a manual compile-installation instruction so that I can add the new kernel to lilo by hand to try it out, knowing the working kernel is still safe. You can still use make-kpkg and have it not touch the symlinks or the bootloader, they're adjustable via kernel-img.conf. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Virus Scanner 64 bits version ??
On 10/18/05 01:58:53AM +0200, Michele Concina wrote: mmh..just a stupid question..do i need a virus scanner? r there virus for linux? i thought that was impossible to be infected coze how can a virus install itself if i use my os not as root? I'm just a newbie so i don't know so much about security on debian and GNU/Linux systems michele Virus scanners on Linux are largely worthless right now since 99% of the definitions they use are for Windows viruses. But don't fool yourself into thinking that Linux is impregnable. The kernel has had it's share of local root exploits and most network daemons have had some kind of remote exploit in their lives. A virus may not be able to infect /bin/ls as your regular user, but it sure could delete all of your personal data and mail itself to your friends. And most people care more about their own data than their OS, I can install Debian sid in a very small amount of time but replacing all of my data would be virtually impossible since I don't back things up like I should =) I'm not recommending that you install a virus scanner, I'm just saying that you shouldn't be any more lax in your security procedures since you're running Linux. Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: mplayer
On 10/02/05 01:48:48PM +0200, hjalmar wrote: Hello, I am having a little bit of trouble with mplayer and wondering if you guys could give me some help. I apt-get install from deb http://spello.sscnet.ucla.edu/marillat/ sid main Everything went fine but I can only play movies if I am root. Here are the two printouts I get when I start mplayer For root mplayer -vo xv movie name [lots of mplayer output snipped] I have checked the /dev/rtc and the permission is crw-rw-r-- 1 root root 10, 135 2005-03-01 00:08 /dev/rtc and I am a member of the video group as well. Run mplayer with the -v switch, it'll print some debug output to give an idea of what it's trying to do. For instance, when running it here, I get this during the mplayer xv startup: X11 opening display: :0.0 vo: X11 color mask: FF (R:FF G:FF00 B:FF) vo: X11 running at 1680x1050 with depth 24 and 32 bpp (:0.0 = local display) [x11] Detected wm supports layers. [x11] Detected wm supports NetWM. [x11] Detected wm supports FULLSCREEN state. [x11] Detected wm supports ABOVE state. [x11] Detected wm supports BELOW state. [x11] Current fstype setting honours LAYER FULLSCREEN ABOVE BELOW X atoms Disabling DPMS DPMSDisable stat: 1 [xv common] Drawing colorkey manually. [xv common] Using colorkey from Xv (0x0101fe). Opening video filter: [lavcdeint] Thanks for any help, Clyde Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: remove unwanted modules
On 09/21/05 12:24:06AM -0500, Marc DM wrote: Stupid questions : How can I find out which modules my system actually needs and disable the ones I don't need. How can I know if a module I'm disabling at startup isn't needed for another module that I plan to load? Thanks. And I won't ask anymore stupid questions for the rest of the week. But why do you want to do this? A full modules directory in /lib/`uname -r` only takes up ~40M. And who knows when you'll plug in some new USB device or something and wish you had that module handy. The only way you're going to get an accurate list is if you know what all of the modules are used for. And stripping out the modules with a 0 reference count won't be enough, for instance ide-cd currently has a ref count of 0 on this machine because there's no disc mounted. Hmmm and somehow tulip has a ref count of 0 as well even though I know I'm using that NIC... Anyway, you'll pretty much have to go through the lsmod output and run modinfo on each module and decide whether it's important enough to keep or not. Regards, Marc DM Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: remove unwanted modules
On 09/21/05 09:06:44PM -0500, Marc DM wrote: Jim Crilly wrote: But why do you want to do this? A full modules directory in /lib/`uname -r` only takes up ~40M. And who knows when you'll plug in some new USB device or something and wish you had that module handy. Actually, I wanted to know just for knowing purposes. The other reason I wanted to know is because I'm using Debian with a single Opteron246 to create a router to handle traffic between 4 vlans and the internet. So I wanted to make sure that I didn't have any modules in there that might be a potential security threat nor any that would degrade performance solely due to its presence. Know of any? Just a guess, but if a module was known to be a security problem it would most likely have been removed or fixed =) And since you need to be root (or at least have CAP_SYS_MODULE) to load/unload modules, the box will already be compromised by the time they can load any potentially malicious modules. And as for performance, I really doubt any modules would slow anything down to the point where you would notice. Most of the modules that might affect performance require you to do something to activate them, like even if you load every iptables module available it won't matter unless you have rules to make them do something. Especially with a box as fast as an Opteron. You might end up with a little less free memory if you load a few modules that you don't plan on using, but most modules are only few K each anyway. Marc DM Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Rescue CD wanted: 2.6.10, SATA and XFS
On 06/10/05 12:09:35PM +0200, Thomas Steffen wrote: On 6/10/05, Ed [EMAIL PROTECTED] wrote: Knoppix meets the requirements. Beware XFS with NFS and/or Samba. I have recently moved all my clients systems off XFS. I agree. As I posted before, I am not happy with XFS either, but I have not made the effort yet to change it. I will probably change the system partition to reiserfs as soon as the system is up again :-) OT, but IMO reiserfs is a huge stepbackwards from XFS... And Knoppix does support SATA now? Unfortunately, my mainbord has no ATA emulation, so I do need the SATA drivers from kernel 2.6.10 or later. Can you recommend a small variant of Knoppix? I do have ADSL, but 700 MB still take some time... Knoppix 3.9 comes with 2.6.11, so if your driver is in 2.6.10 you should be fine. If you use BitTorrent the download should fly, according to the tracker there's 400 seeds on the EN version. Thomas Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Rescue CD wanted: 2.6.10, SATA and XFS
On 06/10/05 10:50:43AM -0400, Wes Williams wrote: I've been very happy with XFS and have migrated TOWARDS XFS all around, save the boot partition. BTW, being able to boot from XFS would be a nice touch! I agree, I'm using XFS on i386, Alpha and Sparc64 without any issues. What problems in particular are you experiencing with XFS and NFS/Samba? I don't use Samba much any more since I don't have any Windows boxes but in the past I never had an issue with it, but I can't speak about NFS. Just a me too message. I've been using XFS since SGI put out their 1.0 Linux port and while there have been issues, I consider it much better than reiserfs. Jim. -Original Message- From: Jim Crilly [mailto:[EMAIL PROTECTED] Sent: Friday, June 10, 2005 10:34 AM To: Thomas Steffen Cc: Ed; debian-amd64@lists.debian.org Subject: Re: Rescue CD wanted: 2.6.10, SATA and XFS On 06/10/05 12:09:35PM +0200, Thomas Steffen wrote: On 6/10/05, Ed [EMAIL PROTECTED] wrote: Knoppix meets the requirements. Beware XFS with NFS and/or Samba. I have recently moved all my clients systems off XFS. I agree. As I posted before, I am not happy with XFS either, but I have not made the effort yet to change it. I will probably change the system partition to reiserfs as soon as the system is up again :-) OT, but IMO reiserfs is a huge stepbackwards from XFS... And Knoppix does support SATA now? Unfortunately, my mainbord has no ATA emulation, so I do need the SATA drivers from kernel 2.6.10 or later. Can you recommend a small variant of Knoppix? I do have ADSL, but 700 MB still take some time... Knoppix 3.9 comes with 2.6.11, so if your driver is in 2.6.10 you should be fine. If you use BitTorrent the download should fly, according to the tracker there's 400 seeds on the EN version. Thomas Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Rescue CD wanted: 2.6.10, SATA and XFS
On 06/10/05 11:23:55AM -0400, David Wood wrote: Sadly, when measuring a filesystem's reputation, if (controlling as best you can for hardware faults) 9 out of 10 people say it works great, never had a problem, while one says it blew up on them, that's a _terrible_ filesystem. True. And that's why I don't use reiserfs, I've had it blow up in my face a few times. The last time was incredibly fun to debug because when the box would crash the screen would blank so that I couldn't see the oops. I had to hook up a serial console just to figure out that it was reiserfs causing the issues and even then reiserfsck and mounting the filesystem both said it was just fine even though it would oops and hang when accessing a certain file. Once I figured out which file, I deleted it, backed up the volume and formatted it XFS and haven't had a problem since. IIRC that was around a year ago so that particular issue might be fixed, but I'm still not a fan of reiserfs or Hans Reiser either. I need to check, but I think, statistically, XFS is faring _worse_ than 9 out of 10, on this list at least. Statistically you might be right, for instance there was a problem in 2.6.8 or 2.6.9 IIRC, that caused problems with XFS even though I don't think it was directly XFS' fault. And somehow I avoided all of those problems, even using risky software like swsusp2. But the only issue I've ever had with XFS is that if it crashes during a write the file being written to may be NULL u filled, but that's pretty rare, at least for me. Jim. On Fri, 10 Jun 2005, Jim Crilly wrote: On 06/10/05 10:50:43AM -0400, Wes Williams wrote: I've been very happy with XFS and have migrated TOWARDS XFS all around, save the boot partition. BTW, being able to boot from XFS would be a nice touch! I agree, I'm using XFS on i386, Alpha and Sparc64 without any issues. What problems in particular are you experiencing with XFS and NFS/Samba? I don't use Samba much any more since I don't have any Windows boxes but in the past I never had an issue with it, but I can't speak about NFS. Just a me too message. I've been using XFS since SGI put out their 1.0 Linux port and while there have been issues, I consider it much better than reiserfs. Jim. -Original Message- From: Jim Crilly [mailto:[EMAIL PROTECTED] Sent: Friday, June 10, 2005 10:34 AM To: Thomas Steffen Cc: Ed; debian-amd64@lists.debian.org Subject: Re: Rescue CD wanted: 2.6.10, SATA and XFS On 06/10/05 12:09:35PM +0200, Thomas Steffen wrote: On 6/10/05, Ed [EMAIL PROTECTED] wrote: Knoppix meets the requirements. Beware XFS with NFS and/or Samba. I have recently moved all my clients systems off XFS. I agree. As I posted before, I am not happy with XFS either, but I have not made the effort yet to change it. I will probably change the system partition to reiserfs as soon as the system is up again :-) OT, but IMO reiserfs is a huge stepbackwards from XFS... And Knoppix does support SATA now? Unfortunately, my mainbord has no ATA emulation, so I do need the SATA drivers from kernel 2.6.10 or later. Can you recommend a small variant of Knoppix? I do have ADSL, but 700 MB still take some time... Knoppix 3.9 comes with 2.6.11, so if your driver is in 2.6.10 you should be fine. If you use BitTorrent the download should fly, according to the tracker there's 400 seeds on the EN version. Thomas Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Gnome broken in Sid?
On 06/11/05 08:48:06AM +1000, Pete wrote: Hi all, I did my first update in a week or so today, and now find that Gnome is doing very strange things indeed. Firstly, gnome-panel appears to be gone. I have no menus at all, only desktop icons. When I try dselect and look through things, most Gnome items appear to be 2.8.2, but gnome-panel is 2.8.3 and gnome-panel-data is 2.10!!! If I try to install gnome-panel, it wants to remove most of my installed applications. I can't seem to find a bug report about this, so not sure if there isn't one or if I'm looking in the wrong place. Can anyone confirm that this is affecting them too or if I'm just going nuts? Yes, the Gnome 2.10 build started earlier this week and parts of it haven't made it into sid yet so things are broken. I don't have a full Gnome installation, so the only thing I noticed is that libgnomevfs2-common is 2.10.1-4 while gnome-panel and gnome-panel-data are at 2.8.3-1 so upgrading libgnomevfs2-common wants to remove them. If you hadn't upgraded yet, I would have recommened that you just put the Gnome stuff on hold for a few more days until Gnome 2.10 finishes it's way into sid. Pete Jim. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
