May 2004 - openSUSE AMD64 - openSUSE Mailing Lists

9.1 & openafs
by Andrej Filipcic 23 May '04

23 May '04

Hi, I have a few questions about 9.1 and openafs. As I understand, there is no stable 2.6 afs client yet. Will there be a release of the client, when it is available? Is it possbile to run 9.1 with 9.0 2.4.21 kernel? Thanks, Andrej _____________________________________________________________ doc. dr. Andrej Filipcic, E-mail: Andrej.Filipcic(a)ijs.si Department of Experimental High Energy Physics - F9 Jozef Stefan Institute, Jamova 39, P.o.Box 3000 SI-1001 Ljubljana, Slovenia Tel.: +386-1-477-3674 Fax: +386-1-425-7074 -------------------------------------------------------------

2 1

Re: [suse-amd64] Suse9.1 still fails to install
by sean-lynch＠sean-lynch.com 23 May '04

23 May '04

only one floppy. The images will be listed in your config file. I know and use Lilo so I will give you an example using Lilo (at the bottom). I am sure there are plenty of examples for grub users to follow. Google is your friend. If you don't want to mess with config files, just install the boot loader from SuSE. Yast will allow you to add as many of your other OS's using a point and click gui interface. I'm sure Redhat or Mandrake has the same ability. At least I assume they do, I just haven't use them in six or seven years. As to the rebooting during install. At some point in the install of every OS you must move from the operating system on the install media to the OS you have just installed on a hard drive. For many years Linux installs worked like SuSE still does. Install enough to get started, reboot early and finish the install and configuration. This was always seen as a practical way to handle installing Linux, because so much had to be configured by hand, and you were going to have to recompile the kernel for your device drivers. A lot has changed Contrast this to the Windows family of OS's. They install almost everything, and reboot very close to the end of installation. right at the end you reboot to your hard disk and there is just a teeny bit of configuration left. There is still a reboot. I know that you can also 'reboot on the fly'. switching from the OS on the install media to the installed OS. You can do it warm and hide it behind a splash screen or something. I don't know if RedHat or Mandrake do this, but theoretically it can be done. Seems like a waste of effort to me, but hey. But there is still a point where you stop using the OS on your install media and start using the os on your hard drive. So there must be some point where you reboot, early middle or end. Unless you use knoppix and always boot from CD and never install to hard drive. # example /etc/lilo.conf boot = /dev/hda # place lilo in MBR of first disk delay = 1000 # Wait 10 seconds, to give time to hit a key default = mandrake # the image labelled 'mandrake' will boot # if there is no intervention. other = /dev/hda1 # a windows partition label = windows # the label can be ny name table = /dev/hda # where this os will boot from image = /boot/zImage-2.4.22 # A Linux kernel, located in the /boot directory root = /dev/hda2 # The partition that will be mounted as root label = mandrake # The name you should type in to boot this kernel. image = /boot/zImage-2.6.10 # A Linux kernel, located in the /boot directory root = /dev/hda6 # The partition that will be mounted as root label = redhat # The name you should type in to boot this kernel. image = /boot/vmlinuz_suse # A Linux kernel, located in the /boot directory root = /dev/hda8 # The partition that will be mounted as root label = suse # The name you should type in to boot this kernel. I'm sure grub has many examples on the internet you can follow. On Fri, 21 May 2004 11:56:25 -0400 (EDT), michael12(a)mindspring.com wrote: > > I'll try, but I have quite a few different boot images. Will I need one floppy > disk per image? Also, neither Redhat nor Mandrake has ever required a reboot at > this stage. It's just insane... > > Michael > > -----Original Message----- > From: sean-lynch(a)sean-lynch.com > Sent: May 21, 2004 10:39 AM > To: suse-amd64(a)suse.com > Subject: Re: [suse-amd64] Suse9.1 still fails to install > > Oh! This is a problem. SuSE still does a reboot after installing the basic > system. The reboot gets you off the dvd booted system and on to your newly > installed system. Then there is more installation and setup after you do this > re-boot. > > If you did not install a bootloader, you will not be able to boot your newly > installed SuSE to finish the install. Historically I believe almost all Linux > ditributions used this reboot half way through install. > > Here are two solutions to try A or B (please read both): > > A) > 1)Boot to Mandrake. Make a mandrake boot floppy. confirm you can boot to > Mandrake with the boot floppy. > > 2)Install SuSE. Install the bootloader to your harddrive (yes, overwriting the > Mandrake boot loader, but don't worry you have a Mandrake boot floppy from point > 1 above!). Completely install SuSE past the mid install re-boot, Sax X setup, > post configuration. create a SuSE boot floppy now. Reboot again to make sure > SuSE is perfectly installed. You might add Mandrake and any other OS's you have > to the SuSE boot loader and quit here, or print out (or write down) the lines > from Grub (or Lilo) config files for the SuSE boot. > > 3)Use your Mandrake boot floppy to boot to Mandrake. Edit your Mandrake Grub (or > Lilo) config file to include SuSE. Update Grub (or Lilo) with these changes and > re-install it to your hard drive. This overwrites your SuSE installed boot > loader, but you get all your old Mandrake settings back, plus a new entry for > SuSE. > > B) > On the other hand you may not have to start completely over... > > You might be able to use the SuSE install DVD as a rescue disk to boot to an OS > on the hard drive. Choose the partion you started your SuSE install to and boot > to it. You might be able to pick up the installation where you left off, even > without a boot loader installed. If this is the case, create a SuSE boot floppy > as soon as you get a working system. > > Good Luck, > > Seï¿½n > > > On Fri, 21 May 2004 08:48:23 -0500 (GMT-05:00), michael12(a)mindspring.com wrote: > > > > > I tried that. It reset before I was able to set up the root password, printer, > > XF86, etc... Basically, just after the RPM installs completed. When I tried > > booting from the DVD again, the DVD started as if it was a new install. Is > > there something else I'm supposed to do? I don't recall anything in the manual > > saying "Your system will reset before you can finish configuring it". I don't > > even have a root password, my XF86 hasn't been set up, etc... Neither Redhat > > nor Mandrake work that way - why does Suse work that way? > > > > Thanks, > > Michael > > > > -----Original Message----- > > From: Andi Kleen <ak(a)suse.de> > > Sent: May 21, 2004 3:44 AM > > To: Michael White <michael12(a)mindspring.com> > > Cc: suse-amd64(a)suse.com > > Subject: Re: [suse-amd64] Suse9.1 still fails to install > > > > > My plan was to add an entry in my Mandrake bootloader for Suse. > > > > > > Well, this crashes too, but in a much worse way. I look down after all the > > > RPMs have been installed, and when I look up again I see Yast starting up. > > > Immediately after that, the system starts shutting down. No real indication > > > as to why this is happening. I didn't see anything of any significance in > > > > SUSE always reboots once during the installation. If you had looked > > you had seen a dialog box telling you that. > > > > This is also documented in the handbook. > > > > Just make it somehow boot again and then the installation will continue. > > > > -Andi > > > > > > -- > > Check the List-Unsubscribe header to unsubscribe > > For additional commands, email: suse-amd64-help(a)suse.com > > -- > Check the List-Unsubscribe header to unsubscribe > For additional commands, email: suse-amd64-help(a)suse.com > > > > -- > Check the List-Unsubscribe header to unsubscribe > For additional commands, email: suse-amd64-help(a)suse.com

1 0

Updating motherboard bios without access to MS Windows
by Hilary L Hertzoff 22 May '04

22 May '04

I have a MSI K8T Neo motherboard on which I'd like to update the bios in the hope that it will stop some periodic nasty crashes that I've been having (both the keyboard and the mouse stop responding, so that my only option is a hard reboot). The instructions on the motherboard site are for Windows only. Is there a process or a utility I can use to do it in linux? I feel like I'm missing something obvious here, but this is new territory for me. I'm currently running SuSE 9.0 Pro. (x86-64). Thanks Hilary

3 4

Win4lin/Nettraverse Kernel Patches + 9.1(amd64)
by Joel Wiramu Pauling 22 May '04

22 May '04

Hi there, Before I go off on a wild mission to get win4lin working (vmware refuses to work with 2.6 kernels and probably won't untill they decide to support it... if they ever do). Can anyone tell me if the default suse kernel is patched for win4lin, and if not anyones experience getting it patched. Kind regards Joel

2 1

How to update kernel from 2.4 to 2.6 on a 4-cpu AMD64 machine with SuSE8.0
by Fan Zhihua 22 May '04

22 May '04

Hi, all! I downloaded a 2.6.5 kernel from the ftp://ftp.kernel.org, and install the module-init-tools. Then I compiled a new kernel and started up, but when I use "top" command, I found only one cpu. When I comiled the pvfs.ko, pvfs could be mounted, but couldn't read or write. When I download the kernel-2.6.4-52-source.rpm of suse9.1 from ftp.suse.com. I find that it needs so many dependencies. After "rpm -i kernel-source-2.6.4-52.src.rpm" and "rpm -bb kernel-source-2.6.4-52.spec --nodeps" and rpm -i the kernel rpm, I can compile the 2.6kernel. I compile the MPT driver into kernel, it can not started up. I try to install the coreutils rpm package etc from ftp.suse.com, but it needs many other packages. Is this way going well? A new guy waiting for your answers. --------------------------------- Do You Yahoo!? 嫌邮箱太小？雅虎电邮自助扩容！

2 1

How to update kernel from 2.4 to 2.6 on a 4-cpu AMD64 machine with SuSE8.0
by Fan Zhihua 22 May '04

22 May '04

Hi, all! I downloaded a 2.6.5 kernel from the ftp://ftp.kernel.org, and install the module-init- tools. Then I compiled a new kernel and started up, but when I used "top" command, I found only one cpu. When I comiled the pvfs.ko, pvfs could be mounted, but couldn't be read or write. Then I downloaded the kernel-source-2.6.4-52.src.rpm of suse9.1 from ftp.suse.com. I find that it needs so many dependencies. After "rpm -i kernel-source-2.6.4-52.src.rpm" and "rpm -bb kernel-source-2.6.4-52.spec -- nodeps" and rpm -i the kernel rpm, I can compile the 2.6kernel. I compile the MPT driver into kernel, it can not started up. I try to install the coreutils rpm package etc from ftp.suse.com, but it needs many other packages also. Is this way going well? A new guy waiting for your answers. --------------------------------- Do You Yahoo!? 嫌邮箱太小？雅虎电邮自助扩容！

1 0

Suse 9.0, 2.6 kernel and hotplug
by Mike Rosing 22 May '04

22 May '04

Howdy folks, I've just ordered 9.1, but it's on back order. I'm hoping that will fix my problem of not getting a thumb drive (or a cd recorder) to work. The 2.6 kernel sees my cd as /dev/hdc, but cdrecord wants to see it as a scsi device. I can read cd's and dvd's just fine, but I can't get cdrecord to work. So I went and got a thumb drive hoping I could just transfer data to it and then use dos (blech!) to burn cd's. Unfortunatly, thumb drives also use the scsi system for creating a virtual disk! After some mucking around I finally got the following in my /var/log/messages: May 20 21:21:34 linux /etc/hotplug/usb.agent[1872]: Bad USB agent invocation May 20 21:21:35 linux kdm[1816]: pam_unix2: session started for user drmike, service xdm May 20 21:21:35 linux resmgr[1117]: accepted connection from user root May 20 21:21:35 linux resmgr[1117]: disconnect from root May 20 21:21:37 linux kernel: Initializing USB Mass Storage driver... May 20 21:21:37 linux kernel: scsi0 : SCSI emulation for USB Mass Storage devices May 20 21:21:37 linux /sbin/hotplug[1954]: no runnable /etc/hotplug/scsi_host.agent is installed May 20 21:21:37 linux kernel: Vendor: Model: Memorex TD 2B Rev: 1.07 May 20 21:21:37 linux kernel: Type: Direct-Access ANSI SCSI revision: 02 May 20 21:21:37 linux /sbin/hotplug[1961]: /etc/hotplug/scsi.agent: ./hotplug.functions: line 41: syntax error near unexpected token `>' May 20 21:21:37 linux /sbin/hotplug[1961]: /etc/hotplug/scsi.agent: ./hotplug.functions: line 41: ` all) exec &> >(logger -t $0[$$]);;' May 20 21:21:37 linux kernel: SCSI device sda: 243712 512-byte hdwr sectors (125 MB) May 20 21:21:37 linux kernel: sda: assuming Write Enabled May 20 21:21:37 linux kernel: sda: assuming drive cache: write through May 20 21:21:37 linux /sbin/hotplug[1972]: no runnable /etc/hotplug/block.agent is installed May 20 21:21:37 linux kernel: sda: sda1 May 20 21:21:37 linux kernel: Attached scsi removable disk sda at scsi0, channel 0, id 0, lun 0 May 20 21:21:37 linux kernel: Attached scsi generic sg0 at scsi0, channel 0, id 0, lun 0, type 0 May 20 21:21:37 linux /sbin/hotplug[1979]: no runnable /etc/hotplug/block.agent is installed May 20 21:21:37 linux /sbin/hotplug[1981]: no runnable /etc/hotplug/scsi_generic.agent is installed May 20 21:21:37 linux kernel: WARNING: USB Mass Storage data integrity not assured May 20 21:21:37 linux kernel: USB Mass Storage device found at 2 May 20 21:21:37 linux kernel: drivers/usb/core/usb.c: registered new driver usb-storage May 20 21:21:37 linux kernel: USB Mass Storage support registered. May 20 21:21:37 linux /sbin/hotplug[1980]: no runnable /etc/hotplug/scsi_device.agent is installed May 20 21:21:38 linux /sbin/hotplug[1961]: disk at /devices/pci0000:00/0000:00:10.2/usb3/3-2/3-2:1.0/host0/0:0:0:0 Which looks like the kernel can see the drive. But when I try to mount it: linux:/sata0 # mount -t vfat /dev/sda /mnt/trek mount: wrong fs type, bad option, bad superblock on /dev/sda, or too many mounted file systems linux:/sata0 # mount -t fat /dev/sda /mnt/trek mount: fs type fat not supported by kernel The thumb drive has been written to by winxp, so it ought to be vfat. The fat fs type _is_ supported by the kernel, I can read floppies just fine. The kernel cold plugs the thumb drive, and reports it's attached to sda1. My fstab entry is: /dev/sda /mnt/trek vfat auto,users 0 0 It seems like I'm almost there, but I can't figure out what I'm missing. Anyone else figured this out, or do I just wait for 9.1 to show up and hope things will work? Any comments appreciated! Patience, persistence, truth, Dr. mike

3 6

RE: [suse-amd64] Re: [discuss] Clock problem on Suse 9.0
by Bedard, Joe 22 May '04

22 May '04

Hi, I'm having a similar problem with my Celestica 4-way box. 2.4.21-212-smp x86_64 SLES8: /proc/interrupts CPU0 CPU1 CPU2 CPU3 0: 278917516 0 0 0 IO-APIC-edge timer 1: 20 0 0 0 IO-APIC-edge keyboard 2: 0 0 0 0 XT-PIC cascade 9: 0 0 0 0 IO-APIC-edge acpi 14: 19695 2 0 0 IO-APIC-edge ide0 26: 167539493 0 0 0 IO-APIC-level eth0 29: 42 0 0 0 IO-APIC-level ioc0 30: 16770775 0 0 0 IO-APIC-level ioc1 33: 294692805 0 0 0 IO-APIC-level cciss0 37: 173146195 0 0 0 IO-APIC-level cciss1 NMI: 131633663 136034955 127714971 123530408 LOC: 278918112 278918109 278918042 278918111 ERR: 0 MIS: 0 Here is a two way Proliant Xeon with HT: 2.4.21-190-smp i686 SLES8: /proc/interrupts CPU0 CPU1 CPU2 CPU3 0: 170 3876269 0 4047780 IO-APIC-edge timer 1: 11 0 0 0 IO-APIC-edge keyboard 2: 0 0 0 0 XT-PIC cascade 8: 2 0 0 0 IO-APIC-edge rtc 9: 0 0 0 0 IO-APIC-edge acpi 14: 33 0 0 0 IO-APIC-edge ide0 30: 3 254888 0 342651 IO-APIC-level eth0 31: 2967 22914 0 36302 IO-APIC-level cciss0 NMI: 0 0 0 0 LOC: 7923576 7923575 7923575 7923574 ERR: 0 MIS: 0 I found other discussion about this issue and it has causes errors in NIC's, HBA's and clocks. The discussions could not determine if it was a Linux issue or a BIOS ACPI/APIC issue.

2 1

Memory problems with root processes
by Dr. Harry Knitter 22 May '04

22 May '04

Hi there, I am running SuSE 9.1 pro on a AMD Athlon64 3000+ 512 MB RAM. There is a strange behaviour of the system when trying to backup system files (DDS3 Streamer HP C1537A on Adaptec 2940U). Using Kdat as root (kdesu kdat), transfer rate decreases and system becomes so slow that it almost cannot be used anymore. I watched the usage of memory with ksysguard. Memory filled up and so did the swap-partition. When the system arrives at about 1 GB of swap usage the system almost freezes. Doing the same as a user backing up user data (almost the same amount of data) there are not any of these problems. The same behaviour can be realized when doing a system backup with Kdar. I assume that these root-processes grab all the memory available. How can I limit memory usage of root processes so that they behave like they do when they are running with ordinary user privileges? Thanks in advance Harry

1 0

Re: [discuss] Clock problem on Suse 9.0
by Milan Gabor 21 May '04

21 May '04

Hi! Yes, it is SMP dual opteron on MSI K8 Master2 board. Kernel 2.4.21-215-smp. CPU0 CPU1 0: 41368 133254241 IO-APIC-edge timer 1: 680 0 IO-APIC-edge keyboard 2: 0 0 XT-PIC cascade 14: 30 1 IO-APIC-edge ide0 16: 3248342 0 IO-APIC-level eth0 20: 3645910 2825 IO-APIC-level libata NMI: 3244855 4761740 LOC: 133274580 133274413 ERR: 1 MIS: 0 Is there any solution to this problem? MIlan Samuel Flory wrote: > Milan Gabor wrote: > >> Hi! >> >> I noticed lately, that system clock on my dual opteron system is going >> slower than real time. In one day clock goes wrong about 10 seconds :( >> >> What could cause this behaviour? >> >> FOr couple of months everything was OK. >> > > > > > Is this a smp box with a high interrupt load? (cat /proc/interrupts)

1 0