On 2007/10/30 08:26 (GMT+0200) Stan Goodman apparently typed:

Specifying the hardware is very easy: This is an Intel board, 915GAV, with everything on board, specifically the video. There are no cards in the slots.

I think I need to bite the bullet, and just make an initial installation. Do you agree?

Because of the Intel video, I'd stick with 10.2 until 11.0 has been released. -- "The basis of our Bill of Rights comes from the teachings we get from Exodus and St. Matthew, from Isaiah and St. Paul. President Harry S. Truman Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

Andreas Jaeger wrote:

Felix Miata writes:

...

On 2007/10/30 08:26 (GMT+0200) Stan Goodman apparently typed:

...

Specifying the hardware is very easy: This is an Intel board, 915GAV, with everything on board, specifically the video. There are no cards in the slots.

I think I need to bite the bullet, and just make an initial installation. Do you agree?

Because of the Intel video, I'd stick with 10.2 until 11.0 has been released.

What problem do you have with the Intel video in 10.3?

I'm curious about that a well, as I was about to upgrade my intel-equipped 10.2 workstation to 10.3 Joe -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On 2007/10/31 13:37 (GMT-0400) Felix Miata apparently typed:

https://bugzilla.novell.com/show_bug.cgi?id=257746 (lots of dupes)

I meant for the above to point to a specific comment quoting Intel: https://bugzilla.novell.com/show_bug.cgi?id=257746#c48 -- "The basis of our Bill of Rights comes from the teachings we get from Exodus and St. Matthew, from Isaiah and St. Paul. President Harry S. Truman Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Tue, 30 Oct 2007 09:02:24 -0400

On 2007/10/30 08:26 (GMT+0200) Stan Goodman apparently typed:

...

Specifying the hardware is very easy: This is an Intel board, 915GAV, with everything on board, specifically the video. There are no cards in the slots.

...

I think I need to bite the bullet, and just make an initial installation. Do you agree?

Because of the Intel video, I'd stick with 10.2 until 11.0 has been released.

Really! That is very discouraging, because I am pretty sure it is to late to for that. I suppose that by the time the upgrade gets to the first reboot, what is on the HD isn't really v10.2 anymore. Why are you saying this? What difficulty is known about v10.3 and the video subsystem on the Intel 915G MB? I now know, by the way, why the upgrade moved GRUBB. It is true that, as I said, I was never asked about the location of the loader. As everyone here except me knows, that question is never asked. One has to know to CHANGE the boot arrangements. If one has been around the block sufficiently, it is probably automatic to do that change. That may be adequate for a new installation, but it seems very clear to me that, in the case of an _UPGRADE_, the default ought to be to leave it the loader where it already is; I don't think that logic can be challenged. I am not sure where it has put the loader. If it is now in the MBR (which I think is the default for new installations), I'll make a new MBR. I'll move GRUBB back to where I want it. All this could not have happened at a worse time for me. -- Stan Goodman Qiryat Tiv'on Israel "Happiness is having a large, loving, caring, close-knit family in another city." -- George Burns -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On 10/30/2007 02:26 PM, Stan Goodman wrote:

So what options are available to me now? To discard what is there and make a New Install?

Just use the rescue system to repair GRUB, reboot and complete the install. I also had a grub failure due to an old mdadm.conf file, but fixing it was not that hard and install picked up where it had left off. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On 10/30/2007 11:05 PM, Stan Goodman wrote:

That means that what I have to do is to get into YaST2 and redirect the loader to where it should be, or equivalently to give a command-line instruction that will do the same thing. I don't know how to do either of those things, and the books I have here don't tell me.

I have never used the Rescue entry into SuSE. I have tried to go into it, and am confronted with a demand for an ID and PW. I tried "root" as ID, but I have no idea at all what it will accept as a PW.

What's a reasonable password for Rescue?

Your root password. You said this was an upgrade didn't you?

How can I find how to select the loader location?

In Yast2 boot loader, or etc/sysconfig Editor. In case you have never done this, you would log into the rescue system, then mount the root partition of the drive, i.e. mount /dev/md0 /mnt (Not sure if these are still needed in 10.3 or not) mount -o bind /proc /mnt/proc mount -o bind /dev /mnt/dev mount -o bind /sys /mnt/sys cd mnt chroot /mnt That will put you into your old system as root. To run Yast2 ncurses mode, run yast. to exit out of your change rooted system, type exit. Then shutdown -r now to reboot. HTH. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On 10/31/2007 08:26 PM, Stan Goodman wrote:

...

In Yast2 boot loader, or etc/sysconfig Editor. In case you have never done this, you would log into the rescue system, then mount the root partition of the drive, i.e. mount /dev/md0 /mnt (Not sure if these are still needed in 10.3 or not) mount -o bind /proc /mnt/proc mount -o bind /dev /mnt/dev mount -o bind /sys /mnt/sys cd mnt chroot /mnt That will put you into your old system as root. To run Yast2 ncurses mode, run yast. to exit out of your change rooted system, type exit. Then shutdown -r now to reboot. HTH.

Many, many thanks. I'm sure it would have taken me a long time to acquire all of that by myself. But logging into Rescue involves getting past the demand for an ID and password; I have tried using "root" and the PW that I have been using as root in normal use, but these don't fly. What does it want?

I just rebooted to double check, and it only prompts for user. It does not prompt for a password. User is root. It is not that hard with the above instructions. Are you saying yours IS asking for a password? The rescue system has worked the same way since I have used SUSE (6.4). I have a hard time believing yours works differently than mine. I am using the DVD x86_64 flavor. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from "Joe Morris (NTM)" on Wed, 31 Oct 2007 21:06:21 +0800

On 10/31/2007 08:26 PM, Stan Goodman wrote:

...

...

In Yast2 boot loader, or etc/sysconfig Editor. In case you have never done this, you would log into the rescue system, then mount the root partition of the drive, i.e. mount /dev/md0 /mnt (Not sure if these are still needed in 10.3 or not) mount -o bind /proc /mnt/proc mount -o bind /dev /mnt/dev mount -o bind /sys /mnt/sys cd mnt chroot /mnt That will put you into your old system as root. To run Yast2 ncurses mode, run yast. to exit out of your change rooted system, type exit. Then shutdown -r now to reboot. HTH.

I know a little more now about the booting sequence of the partially upgraded 10.2 > 10.3. GRUB seems still OK to be, it points to a stage-2 loader which is at sector 0x1973e28 (which is inside /dev/sda6 somewhere). (GRUB, I should add occupies its own small (7MB) EXT2 partition, sda7. The reason it is there rather one of the conventional locations is that openSuSE itself is on an XFS file system. XFS has its own peculiarities, and this is a workaround.) GRUB can also read that sector, and the ones directly after it, and they contain valid GRUB datastructures as shown by the level of detail displayed by DFSee (see below). ++++++++++ GRUB stageX @PSN : 0x01973e28 Located on : /dev/hda6 StageX segm:addr : 0800:8000 Forcing LBA : No Stage1 install by : 3.2 Boot Disk # : 0xff = disknr : 1 StageX install by : 3.2 LoadSegment : 0x0820 = stage 2 code GRUB code version : 0.97 Code size : 100.5 KiB GRUB Config file : /boot/grub/menu.lst ++++++++++ So, all is apparently as it should be, and I can't tell why it does not boot ... Perhaps some change (kernel name/revision ?) is needed in the /boot/grub/menu.lst file. Does this ring a bell with anyone? -- Stan Goodman Qiryat Tiv'on Israel Are people more violently opposed to fur than to leather because it's safer to harrass rich women than biker gangs? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Thu, 01 Nov 2007 09:17:42 -0400

On 2007/11/01 14:55 (GMT+0200) Stan Goodman apparently typed:

...

GRUB seems still OK to be, it points to a stage-2 loader which is at sector 0x1973e28 (which is inside /dev/sda6 somewhere).

...

(GRUB, I should add occupies its own small (7MB) EXT2 partition, sda7.

...

GRUB stageX @PSN : 0x01973e28 Located on : /dev/hda6

...

So, all is apparently as it should be, and I can't tell why it does not boot ...

It's above my understanding why Grub needs to be in two partitions, so maybe a DFSee 'part -b-' report would be useful.

There you have me, I didn't know about being in two places. I also don't know about a "part -b report". Do you mean the DFSDISK report? I can send it as an attachment easily. -- Stan Goodman Qiryat Tiv'on Israel "I think there is a world market for about five computers" -- Thomas J. Watson, IBM Chairman, in 1943 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Thu, 01 Nov 2007 11:58:40 -0400

On 2007/11/01 17:06 (GMT+0200) Stan Goodman apparently typed:

...

from Felix Miata on Thu, 01 Nov 2007 09:17:42 -0400

...

...

On 2007/11/01 14:55 (GMT+0200) Stan Goodman apparently typed:

...

...

...

GRUB seems still OK to be, it points to a stage-2 loader which is at sector 0x1973e28 (which is inside /dev/sda6 somewhere).

...

...

...

(GRUB, I should add occupies its own small (7MB) EXT2 partition, sda7.

...

...

...

GRUB stageX @PSN : 0x01973e28 Located on : /dev/hda6

...

...

...

So, all is apparently as it should be, and I can't tell why it does not boot

...

...

It's above my understanding why Grub needs to be in two partitions, so maybe a DFSee 'part -b-' report would be useful.

...

There you have me, I didn't know about being in two places. I also don't know about a "part -b report". Do you mean the DFSDISK report? I can send it as an attachment easily.

The references to both /dev/sda7 and /dev/hda6 and sda7 would seem to imply there are at least two disks on which at least portions of Grub might be found, or maybe one disk with multiple device names.

I'm sorry, that is the result of my clumsiness. There is only one HD, and it is a SATA drive. The DFSee report came back naming "hda" in both places above, and I replaced one of the instances with "sda" as I was mulling over the significance of what appears to be an error of some sort. I forgot to change it back when I sent my note to this forum. The only "hda" on the machine is the CD/DVD drive. That is a certain, as you can see from the fact that, when I need to unmount the CD/DVD drive in order to eject a disk, my command is "umount /dev/hda".

The onscreen report from simply starting DFSee ends with a chart listing partitions and their characteristics. The one you get from running DFSee in Linux instead of DOS, doz or OS/2 is a bit different. From Linux, on the command line after starting DFSee you run 'part -b-' to get a chart identical to the one provided by the other varieties by default. It's essentially an amplified version of output from 'fdisk -l <diskdevice>' for all accessible disks.

I ran DFSee v9.xx from OS/2, because that is the only way I can do it. The whole problem, of course, is that I can't access Linux. I do not have a DFSee v9.xx bootable disk, but I do have one for v8.02, but I am sure this earlier version does not give a much detail as the new v9.xx does. Do you want me to run DFSDISK with DFSee v8? Or some other report? Nothing would be easier, just say the word. I have an Ubuntu disk, which I think can run live. I have never used it for anything. I am assuming that if I boot from it, I could run the Linux DFSee v8 -- but I also assume that I would need a second CD/DVD drive to do this. But could I boot the openSuSE DVD to "Rescue", then remove the DVD and insert the DFSee DVD to run the utility? If so, which partitions would I need to mount, if any? -- Stan Goodman Qiryat Tiv'on Israel REAL similes/metaphors by high school students; #18: Shots rang out, as shots are wont to do. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Thu, 01 Nov 2007 13:00:26 -0400

On 2007/11/01 18:35 (GMT+0200) Stan Goodman apparently typed:

...

I'm sorry, that is the result of my clumsiness. There is only one HD, and it is a SATA drive. The DFSee report came back naming "hda" in both places above, and I replaced one of the instances with "sda" as I was mulling over the significance of what appears to be an error of some sort. I forgot to change it back when I sent my note to this forum.

...

The only "hda" on the machine is the CD/DVD drive. That is a certain, as you can see from the fact that, when I need to unmount the CD/DVD drive in order to eject a disk, my command is "umount /dev/hda".

...

I ran DFSee v9.xx from OS/2, because that is the only way I can do it. The whole problem, of course, is that I can't access Linux. I do not have a DFSee v9.xx bootable disk, but I do have one for v8.02, but I am sure this earlier version does not give a much detail as the new v9.xx does. Do you want me to run DFSDISK with DFSee v8? Or some other report? Nothing would be easier, just say the word.

http://mrmazda.no-ip.com/tmp/dfsee/p-t18f.txt shows you what I was looking for. Run from OS/2, it's simply the content of dfswork.log that DFSee tries to generate when you start it, press <RETURN>, and then exit, except that I've stripped away portions at the head of the file that are not useful for our purposes here. The Linux version does the same thing, except that the format of the table differs somewhat. 'part -b-' would simply append the DOS/OS2/doz-style table to the rest of it.

As long as you're appending files, include /etc/grub.conf. Better yet, attach nothing, but instead post dfswork.log as dfswork1.txt in web space as I did above, and, also as above, put a link to it in your reply.

I do not know how you got the log file to get all that detail. The one I have is much simpler, even allowing for the fact that my hardware is much simplet than yours. Given its size, I think it's OK to just post it here. Here is the file: ***** Appending to log : 'H:\dfswork.log' (7-bit ASCII) Registration : none RgKey status : KEY not present! 22 days left for evaluation! ReadWrite Store 1 : PhysDisk : 1 PartId 00 = -- mode=FDISK 011107-193452 RC:0 <Enter> : 00000000 u=00000000 x=00000000 this=00000000 Base=00000000 ________________________________________________________________________________ 2 ________________________________________________________________________________ 1 Registration : none RgKey status : KEY not present! 22 days left for evaluation! Please install the registration 'DFSee.key' in a directory in the PATH. To get a registration, follow the registration links from: http://www.dfsee.com/dfsee/buynow NO VALID REGISTRATION KEY FOUND! Without a key, D:\DFSEE9\OS2\DFSOS2.EXE should be used for evaluation only! Use the <Enter> key or the mouse to confirm the message. UI TxWindows : 2.02 04-10-2007 (c) 1995-2007: Fsys Software OS version : OS/2 4.50 OS2kernel: 14.104a_W4 on drive C: ________________________________________________________________________________ 1 DFSee OS/2 9.02 : executing: map -M -m Command timestamp : Thursday 2007-11-01 19:34:54 Disk DFSee Disk Image filename or id type nr (Unix) device, and optional partitionable media description ==== ===== ==== =========================================================== 1 Phys 1 /dev/hda Physical disk 1 (IDE/SCSI/USB or PCCARD) 2 Phys 2 /dev/hdb Physical disk 2 (IDE/SCSI/USB or PCCARD) Number of disks : 2 P-Geo Disk 1 Cyl : 16383 H: 16 S:63 Bps:512 Size : 0x09962B80 = 78533.4 MiB Geometry D1 from : LVM info (DLAT) sector at 0x3e L-Geo Disk 1 Cyl : 10011 H:255 S:63 Bps:512 Size : 0x0996055B = 78528.7 MiB BIOS Int13 limit : 1024, I13X support needed beyond : 8032.5 MiB MBR crc 054b4eb9 : 0x0c8ca699 = DFSee generic MBR, English messages, I13X L-Geo Disk 2 Cyl : 512 H: 12 S:32 Bps:512 Size : 0x00030000 = 96.0 MiB DUMMY, disk 2 seems reserved with no media present (USB/PCCARD etc) +-----<[ D1 ]>-------------------------------------------+ | |+===+=======================================================================| | || |+-----++-------------++---+| |+------++------++-------++------+| |m||1 ||2 ||3 ||4 ||5 ||5 ||6 ||7 ||8 || |b|| || || || || ||C: ||D: ||H: ||W: || |r||BMG||SWAP ||XFS ||EXT||FreeSpace||HPFS ||JFS ||JFS ||JFS || | || |+-----++-------------++---+| |+------++------++-------++------+| | |+===+=======================================================================| +------------------------------------------------------------------------------+ ________________________________________________________________________________ 1 DFSee OS/2 9.02 : executing: fdisk -r- Command timestamp : Thursday 2007-11-01 19:34:54 EBR02= warning : A windows-compatible extended-partition type 0x0f is used, Old operating systems might not see all logical partitions. +---+--+-----------------+--+--------+--------+-----------+----------+--------+ |ID |Dr|Type, description|ux|Format |Related |VolumeLabel|LVM Volume|Size MiB| +---+----------+--------+-----------<[ D1 ] >--------+ |01>| |Prim 0a IBM-BMGR | 2|BMGR |LVM |I13Xneeded |., BootMan| 7.8| |02 | |Log 82 SunS/SWAP| 5|SWAP |LinuxV1 |SWAPSPACE2 |, LinuxSwa| 502.0| |03 | |Log 83 LinuxNatv| 6|XFS |Linux | |SuSE, SuSE| 50007.0| |04*| |Log 83 LinuxNatv| 7|EXT2 |GRUB | |SuSE v10.2| 7.8| |05 | |FreeSpace Logical| |-- -- --|-- -- --|- - - - - -| | 21987.4| |05*|C:|Log 07 Inst-FSys| 8|HPFS |IBM 4.50|ECS |eCS v1.1, | 1004.0| |06 |D:|Log 35 Warp-LVM | 9|JFS |IBM 4.50| |OS/2 Apps,| 1506.1| |07 |H:|Log 35 Warp-LVM |10|JFS |IBM 4.50|INFO |Info, Info| 2502.3| |08 |W:|Log 35 Warp-LVM |11|JFS |IBM 4.50|DATAFILES |DataFiles,| 1004.0| +---+--+-----------------+--+--------+--------+-----------+----------+--------+

= Active/Startable * = Bootable r = Removable R = Removable+Active/Bootable

________________________________________________________________________________ 1 Despite what the heading says, I HAVE registered it, but not yet put the key file in -- I am leaving the country in three days, and am panicked with things left to do, which is why this could not have happened at a worse time. Jan has explained to me that DFSee calls ALL disks e.g. "hda" rather than "sda", because it doesn't see the interface anyway. I did not ask him what happens in a system with both serial and parallel HDs. If I have misunderstood your instructions, please give me more detailed ones, and I will follow them as well as I can. -- Stan Goodman Qiryat Tiv'on Israel I hate puns. Except synonym puns, which are good with coffee for breakfast. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Thu, 01 Nov 2007 14:39:48 -0400

On 2007/11/01 19:52 (GMT+0200) Stan Goodman apparently typed:

...

I do not know how you got the log file to get all that detail. The one I have

All that detail is because I have more partitions than you. Otherwise, they're the same output, with mine reduced somewhat by manual editing.

That's what I meant by "my hardware is much simpler than yours".

...

is much simpler, even allowing for the fact that my hardware is much simplet than yours. Given its size, I think it's OK to just post it here. Here is the file:

...

P-Geo Disk 1 Cyl : 16383 H: 16 S:63 Bps:512 Size : 0x09962B80 = 78533.4 MiB Geometry D1 from : LVM info (DLAT) sector at 0x3e L-Geo Disk 1 Cyl : 10011 H:255 S:63 Bps:512 Size : 0x0996055B = 78528.7 MiB BIOS Int13 limit : 1024, I13X support needed beyond : 8032.5 MiB MBR crc 054b4eb9 : 0x0c8ca699 = DFSee generic MBR, English messages, I13X

...

DFSee OS/2 9.02 : executing: fdisk -r- +---+--+-----------------+--+--------+--------+-----------+----------+--------+ |ID |Dr|Type, description|ux|Format |Related |VolumeLabel|LVM Volume|Size MiB| +---+----------+--------+-----------<[ D1 ] >--------+ |01>| |Prim 0a IBM-BMGR | 2|BMGR |LVM |I13Xneeded |., BootMan| 7.8| |02 | |Log 82 SunS/SWAP| 5|SWAP |LinuxV1 |SWAPSPACE2 |, LinuxSwa| 502.0| |03 | |Log 83 LinuxNatv| 6|XFS |Linux | |SuSE, SuSE| 50007.0| |04*| |Log 83 LinuxNatv| 7|EXT2 |GRUB | |SuSE v10.2| 7.8| |05 | |FreeSpace Logical| |-- -- --|-- -- --|- - - - - -| | 21987.4| |05*|C:|Log 07 Inst-FSys| 8|HPFS |IBM 4.50|ECS |eCS v1.1, | 1004.0| |06 |D:|Log 35 Warp-LVM | 9|JFS |IBM 4.50| |OS/2 Apps,| 1506.1| |07 |H:|Log 35 Warp-LVM |10|JFS |IBM 4.50|INFO |Info, Info| 2502.3| |08 |W:|Log 35 Warp-LVM |11|JFS |IBM 4.50|DATAFILES |DataFiles,| 1004.0| +---+--+-----------------+--+--------+--------+-----------+----------+--------+

...

Despite what the heading says, I HAVE registered it, but not yet put the key file in -- I am leaving the country in three days, and am panicked with things left to do, which is why this could not have happened at a worse time.

...

Jan has explained to me that DFSee calls ALL disks e.g. "hda" rather than "sda", because it doesn't see the interface anyway. I did not ask him what happens in a system with both serial and parallel HDs.

When run from Linux, DFSee can easily tell the difference between PATA and SCSI, but I'm not sure about telling the difference between SATA & SCSI.

...

If I have misunderstood your instructions, please give me more detailed ones, and I will follow them as well as I can.

This was close enough, except that you didn't include /etc/grub.conf.

That's because I didn't know how to get it, since I have no access to SuSE. I am very new in Linux, as you have surely detected. In the interest of saving time, which is at a premium right now, if I can retrieve a few things that are not backed up, I think it might be best to just reinstall v10..3 from scratch. I have a Ubuntu v6.06LTS disk, and as I understand it I can boot with it, and gain access to the /home directory where those small things reside. Or I could do the same with a Rescue boot from the 10.3 DVD. Can I impose on you, or Joe, or anybody else, to let me run past you the actions I need to do to accomplish the retrieval of these files? Joe Morris gave me some hints about using Rescue for a different purpose: *****

...

mount the root partition of the drive, i.e. mount /dev/md0 /mnt (Not sure if these are still needed in 10.3 or not) mount -o bind /proc /mnt/proc mount -o bind /dev /mnt/dev mount -o bind /sys /mnt/sys cd mnt chroot /mnt

That was for using YaST. Which of the above "mount" lines are necessary for the different purpose I have now? Now I want to access the /home directory and also a JFS partition that I use for transfer between SuSE and OS/2. In normal SuSE operation, that partition is, "/mnt/transfer", so I would add the line "mount -o bind /mnt /mnt/transfer". Have I generalized correctly from Joe's notes? I would also add a line like "mount -o bind /home /mnt/home", which would cover any subdirectories. I would copy the needed files. Being root, I would not have to deal with permissions. Is all that correct? so I would add another line to the above as follows: Thank you for the advice below. I will apply these hints when I reinstall the OS.

I don't know what went wrong with the install, or what is wrong now. What first I'd do if it was here would be to delete ID 4 and immediately recreate it as 200MiB instead of 7.8MiB. Kernels and initrds have gotten so large that such small /boot partitions are not safe at upgrade time, or certainly if wanting multiple kernels and/or initrds to be available. The RedHat/Fedora installer will scream and holler at so small a boot partition, recommending at least 75M be allocated to it. I used to allocate 78M for /boot, but now do 200M.

Anyway, after recreating it, you need to rescue boot or boot a Linux live CD and run resize2fs on /dev/sda7 to enable access to all of its new size. After doing that, I'd try a rescue boot mode attempt to reinstall grub, either --batch using the /etc/grub.conf file, or manually from the grub prompt. If that wouldn't work, I'd do a reinstall.

Because you have so much freespace, it would be prudent to use some for a new ext3 partition to use from a rescue boot to copy all of /home. That way you could do an install from scratch of either 10.2 or 10.3. Ext3 is accessible via an OS/2 boot by using the ext2 driver from Hobbes. AFAIK, neither XFS nor ReiserFS have OS/2 drivers, and certainly Linux LVM does not.

If I was starting nearly all over, I'd first back up /home as above, then delete all linux partitions except the /home backup. Then I'd make the logical first Linux partition a *primary* ext2 of 200M, next a logical swap (sda5), then a much smaller root (sda6, minimum 4G, up to maybe 10G, more if you're a developer), then permanent /home (sda7).

I'm not sold on the idea that any other type is better than ext3 for the average user.

Also because you have so much unallocated freespace, you could easily allocate 4-10G for another SUSE installation, which could be used as a rescue system if nothing else. On future upgrades it could be a fallback when installation malfunctions. Or like many do, always have two. Use the oldest as the fallback reserve. When a new release comes out, upgrade the older, using the previous newer as the fallback. They're usually a lot easier than "rescue" boots from CD/DVD. -- " A patriot without religion . . . is as great a paradox, as an honest man without the fear of God." John Adams

Team OS/2 ** Reg. Linux User #211409

Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

-- Stan Goodman Qiryat Tiv'on Israel "I did not attend his funeral, but I did send a note of approval" -- Mark Twain, on hearing of the death of a particularly corrupt politician -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from Felix Miata on Thu, 01 Nov 2007 17:50:48 -0400

On 2007/11/01 23:06 (GMT+0200) Stan Goodman apparently typed:

...

In the interest of saving time, which is at a premium right now, if I can retrieve a few things that are not backed up, I think it might be best to just reinstall v10..3 from scratch. I have a Ubuntu v6.06LTS disk, and as I understand it I can boot with it, and gain access to the /home directory where those small things reside. Or I could do the same with a Rescue boot from the 10.3 DVD. Can I impose on you, or Joe, or anybody else, to let me run past you the actions I need to do to accomplish the retrieval of these files?

...

Joe Morris gave me some hints about using Rescue for a different purpose:

...

...

...

mount the root partition of the drive, i.e. mount /dev/md0 /mnt (Not sure if these are still needed in 10.3 or not) mount -o bind /proc /mnt/proc mount -o bind /dev /mnt/dev mount -o bind /sys /mnt/sys cd mnt chroot /mnt

...

That was for using YaST. Which of the above "mount" lines are necessary for the different purpose I have now?

/dev/md0 above does not apply to you. That's his root, corresponding to your /dev/sda6. mnt is an arbitrary mount point. You may choose anything you like, but if what you choose as a mount point doesn't already exist, you must first create it with mkdir. Otherwise, that's a special series of commands prerequisite to an attempt to reinstall Grub or otherwise succeed at a complex rescue boot. For simply retrieving a few files most are not necessary. You might need to run them all if you want to run mc for managing the files you want to retrieve. I suspect the rescue media may not have mc.

...

Now I want to access the /home directory and also a JFS partition that I use for transfer between SuSE and OS/2. In normal SuSE operation, that partition is, "/mnt/transfer", so I would add the line "mount -o bind /mnt /mnt/transfer". Have I generalized correctly from Joe's notes?

I didn't know it was possible to access OS/2 JFS partitions from a Linux boot, so can't help with that approach. However, there's nothing to stop you from using some of that abundance of freespace for a temporary partition that either could access, either FAT32 or HPFS or ext3. From SUSE you can read and write to HPFS, and I think this is also possible from Ubuntu.

Once booted to Linux, whether SUSE rescue or Ubuntu live, you need to mount the partition that has /home on it, and also mount the partition onto which you wish to transfer files from /home.

To make it easiest, try doing it this way once you have completed a root login, but before doing anything else after finishing a Linux boot:

mkdir /source #location to mount source partition mount -t xfs /dev/sda6 /source #mount it mkdir /target #location to mount target partition mount -t hpfs -o rw /dev/sda8 /target #or whatever type and temporary location if you chose to make one first cp -a /source/home/<loginname>/<sourcefiles> /target #puts files you want into the root directory on LVM C: partition

Everything after eash # above is commentary, not to be typed.

If you don't know exactly where the files you want are located under /source/home/<loginname>, remember that 'ls -l | less' acts pretty much like 'dir /p' in OS/2 but with PgUp/PgDn available to facilitate long list examination.

Copying the source files would be easier from a true SUSE rescue (running from /dev/sda6) because you could run Midnight Commander (mc, works very much like FC/2, Larsen Commander and other OFMs available on OS/2). MC isn't on any Ubuntu CD.

If you want to first (before a Linux boot) create a special partition for the temporary transfer, use DFSee to make whatever type, size and location (I'd put it at end of current freespace). Easiest is probably FAT32, since DFSee9 can also format it. Otherwise you'd have to boot whichever OS it is native to and create a filesystem on it (FORMAT in OS/2, mkfs* in Linux).

Many thanks. It's past midnight here, and I am getting bleary, so I may have to ask other questions tomorrow. But I get the drift, and will try to do the retrieval tomorrow (the last day I will have to do anything in this direction). Then I will reinstall from scratch after my return from abroad. As for JFS, apparently the IBM version is compatible with that of Linux, because I have been accessing this partition quite happily from SuSE since July, which was a pleasant surprise.

From what I grasp through my descending eyelids, it seems more promising to go with Rescue than with Live Ubuntu.

Did you receive the message I sent you to your address, about Wankel engines? -- Stan Goodman Qiryat Tiv'on Israel

From REAL answers on children's science examinations (#1): Q: Name the four seasons. A: Salt, pepper, mustard and vinegar. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from "Rajko M." on Thu, 1 Nov 2007 23:53:06 -0500

On Thursday 01 November 2007 12:52:53 pm Stan Goodman wrote:

+---+--+-----------------+--+--------+--------+-----------+----------+--------+ |ID |Dr|Type, description|ux|Format |Related |VolumeLabel|LVM Volume|Size MiB| +---+----------+--------+-----------<[ D1 ] >--------+ |01>| |Prim 0a IBM-BMGR | 2|BMGR |LVM |I13Xneeded |., BootMan| 7.8| |02 | |Log 82 SunS/SWAP| 5|SWAP |LinuxV1 |SWAPSPACE2 |, LinuxSwa| 502.0| |03 | |Log 83 LinuxNatv| 6|XFS |Linux | |SuSE, SuSE| 50007.0| |04*| |Log 83 LinuxNatv| 7|EXT2 |GRUB | |SuSE v10.2| 7.8| |05 | |FreeSpace Logical| |-- -- --|-- -- --|- - - - - -| | 21987.4| |05*|C:|Log 07 Inst-FSys| 8|HPFS |IBM 4.50|ECS |eCS v1.1, | 1004.0| |06 |D:|Log 35 Warp-LVM | 9|JFS |IBM 4.50| |OS/2 Apps,| 1506.1| |07 |H:|Log 35 Warp-LVM |10|JFS |IBM 4.50|INFO |Info, Info| 2502.3| |08 |W:|Log 35 Warp-LVM |11|JFS |IBM 4.50|DATAFILES |DataFiles,| 1004.0| +---+--+-----------------+--+--------+--------+-----------+----------+--------+

...

= Active/Startable * = Bootable r = Removable R = Removable+Active/Bootable

There is another option. You can resize partition ID 4 to some 10 GB, format with ext3 and install 10.3 as a rescue system where you will have all tools available. Boot loader should be installed in boot sector of partition like current.

Yet another option is to use grub prompt that you have: grub> to repair boot configuration.

Some tips can be picked up from: http://en.opensuse.org/SDB:The_Boot_Manager_Grub and http://en.opensuse.org/GRUB has few usefull links.

The Ubuntu disk can be used to mount /dev/sda6 (ID 03) and see what is installed in /boot and /boot/grub directories. If it works like Knoppix than partition should be already mounted as read only and presented as icons on desktop.

It's very good to have options; thanks for all this information. I am very grateful for the helpfulness of this group. -- Stan Goodman Qiryat Tiv'on Israel REAL similes/metaphors by high school students; #13: The hailstones leaped from the pavement, just like maggots when you fry them in hot grease. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from "Rajko M." on Fri, 2 Nov 2007 18:54:03 -0500

On Friday 02 November 2007 01:36:35 am Stan Goodman wrote:

...

It's very good to have options; thanks for all this information. I am very grateful for the helpfulness of this group.

I guess it is time for trip. I hope to hear from you when you are back.

And you will. Now I have to unsubscribe from this group, so that my mailbox doesn't become inundated in my absence. -- Stan Goodman Qiryat Tiv'on Israel Old Fart n. slang (old fart'; in New England, old faaht') Tribal Elder. Used in deprecation especially by males younger than 20 years. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

** Reply to message from "Joe Morris (NTM)" on Tue, 30 Oct 2007 17:15:56 +0800

On 10/30/2007 02:26 PM, Stan Goodman wrote:

...

So what options are available to me now? To discard what is there and make a New Install?

Just use the rescue system to repair GRUB, reboot and complete the install. I also had a grub failure due to an old mdadm.conf file, but fixing it was not that hard and install picked up where it had left off.

It is clear that the MBR has not been disturbed, because the BootManager, which resides in the MBR, is intact. Since GRUBB resides in its own little partition, I am assuming that it too is intact, or at least its partition is still available to have GRUBB installed on it. That means that what I have to do is to get into YaST2 and redirect the loader to where it should be, or equivalently to give a command-line instruction that will do the same thing. I don't know how to do either of those things, and the books I have here don't tell me. I have never used the Rescue entry into SuSE. I have tried to go into it, and am confronted with a demand for an ID and PW. I tried "root" as ID, but I have no idea at all what it will accept as a PW. What's a reasonable password for Rescue? How can I find how to select the loader location? I have never used the

-- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64

-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

-- Stan Goodman Qiryat Tiv'on Israel "It is the height of absurdity to sow little but weeds in the first half of one's lifetime and expect to harvest a valuable crop in the second half." -- Percy Johnston -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org