How to migrate XP, Vista, Linux, BSD and Solaris to a bigger hard disk - Page 10


Page 10 of 12 FirstFirst ... 6789101112 LastLast
Results 136 to 150 of 180

Thread: How to migrate XP, Vista, Linux, BSD and Solaris to a bigger hard disk

  1. #136
    Join Date
    Jan 2010
    Location
    Pennsylvania
    Posts
    5
    In regards to what I have coming up this weekend with a clone; I am working on an older desktop which has a bad hard drive and I am ultimately looking to clone it to a new hard drive that is on the way. Now looking back on page 5 of this topic I see it mentioned that it will actually transfer the bad blocks of data over as well since it is doing a bit to bit copy so what is my best approach for cloning this old drive to the new one somewhat error free.

    Right now I have had GRC SpinRite running on it for 68 hours and it still has 256 hours remaining (it's only 18% complete). But what measures should I take afterwards for best cloning results. Should I look into "ddrescue" as mentioned on page 5 and then run dd with the "noerror" switch?

    Any direction would much appreciated. Thanks in advance.

  2. #137
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13

    With apologies aforehand...

    For saikee...
    I've avoided this inquiry in hopes that I might either get this (frikkin') boot on or run out of ideas. Well, that happened.
    I've read all 5 of your major instructionals in hope that I might avoid bothering you with yet another uninformed inquiry, but...
    So before getting in too deep it's important to know that I will be happy with a simple reference to an applicable post I might have missed.
    And to admit that I may have broken a rule or two in the get go.

    Box is a 2002 Compaq/HP running XPSp3. Original partitioning on the original included a C: with the usual complement running on ntfs. And a D: (Presario-RP) with restore/repair/renew capabilities on Fat32. Pretty standard I think.

    Following what I thought might be an allowable and also operational dd which worked amazingly well (great technique and advice!) and exactly as predicted. It was a lovely dd.
    The D: was put on a FAT32 partition in the beginning of the recipient drive and the C: onto the 3rd primary in ntfs. The D: wasn't helping boot (it just wanted to wipe the whole recipient disc) so I took that out but left the FAT32 format as is otherwise.
    I worried that the D: was in a first primary (where? windows likes to boot) and the C: was in the third...separated by an extended for the two Linux)

    (As you will see from the Parted Magic image I've included or the table from the same source)

    The Linux OSs I'm running on the recipient disc (Kubuntu & Mepis) BOTH
    list the Windows partition correctly in their respective GRUB menus.
    And both will access and manipulate files in the unbootable M$ partition..
    all the data seems entirely ok.

    I've tried changing the boot flag, getting Mepis to put GRUB at root (which it seems to refuse to do?..I don't know how to 'make' my distros do as you've suggested. I also don't know how to use the Kubuntu 9 to tweak the GRUB location. It no longer has a GRUB application evident in 'system'. Both /boot/grub/menu.lst(s) seem o.k.

    Tried several boot codes (mostly just uninformed mimicry from codes I've seen or made up on the fly or) using the SuperGrubDisk and its GRUB command. line (which is also unable to get Windows up).
    SGD is otherwise very useful...if I corrupt (usually the Mepis) GRUB.

    The original M$ disc is fine..has never been alive with the recipient disc
    from the beginning. I just switch the plugs back and forth when I decide to.
    Pain.

    So this is getting long and I guess my questions are:
    -Any sorta easy fix..I'm not so great on scripting or compiling unless I can straight copy it? Like a fresh boot code from the information I've forwarded
    -Can I just switch partions around...if its just a matter of the order of things
    -Or should I best just get another clean disc and so do - as you've advised in your B: instructional - Whole disc t=>o whole disc. Regardless of the formats thereupon.
    -Or any other[IMG]/home/Owner1/Desktop/PNG image - 839X504 Pixels[/IMG]

    thanks very, sorry to be a trouble
    P.S. Can't seem to attach the Ksnapshot .png of the partitions screenshot/ Parted Magic.
    Will e-mail if it'll help
    Attached Files Attached Files

  3. #138
    Join Date
    Jun 2004
    Location
    Newcastle upon Tyne
    Posts
    2,978
    You seem to move the Windoze partition around and this is not adisable. Every Xp has a hidden file called boot.ini staying which partition the system was originally installed and it is better not to temper with it unless you know exactly where you are doing.

    In your case I would advise a direct disk to disk cloning which is what my thread recommends. In a disk to disk cloning you copy sectors, 512 bytes per sector, of information across. The first sector is the MBR with the partition table inside. Therefore the recipient disk will have exactly the same partition as the source and everything will work once you remove the source and boot the recipient disk alone.

    Only after you have satisfied the recipient disk is working perfectly then you should resize the partitions if required. The partition position of the Xp should never be altered as it brings more trouble than benefit. This is also recommended for the utility parition.

    I suggest to redo the cloning and get the system transferred in working order first.
    Linux user started Jun 2004 - No. 361921
    Using a Linux live CD to clone XP
    To install Linux and keep Windows MBR untouched
    Adding extra Linux & Doing it in a lazy way
    A Grub menu booting 100+ systems & A "Howto" to install and boot 145 systems
    Just cloning tips Just booting tips A collection of booting tips

    Judge asked Linux "You are being charged murdering Windoze by stabbing its heart with a weapon, what was it?" Replied Linux "A Live CD"

  4. #139
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13

    A thing DONE, then....

    ..with my very sincere thanks. And to JUSTLINUX,of course, as well.

    "But I had this nice unused partition and it was a really nice dump.
    Which seemed to work so nice except for the boot part
    And I'm really gonna be working hard to see that
    I never have to buy another M$ OS...ever again
    [I've bought like...4 or...5(?) an' a fella still can't get no break],
    Except that there a couple of applications which are
    really bonded with Windoze and...."
    And so and thusly he mumbly maundered on. And on.

    But you know all that.

    Very much appreciate all your great research and this very prompt reply.
    Been wanting to work into an SSD anyway.
    And at my age a big unbreakable drive will last a lifetime ..or more.

    cheers

  5. #140
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13

    saikee -- Just an hypothetical and...

    ...having abandoned that fruitless effort.

    -IF I had dd'd the C: into a capacious ntfs partition to follow immediately the D: (utility) partition
    which I did put first in the recipient disc,
    which did clone functionally in that FAT32 pre-formatted partition?

    -WOULD that more likely have worked for a good M$ boot?
    Since that would have replicated the original order & placement configuration.

    There would still have been other stuff in the following partitions.

    [I just read your nice reply post more carefully and wondered]

  6. #141
    Join Date
    Jun 2004
    Location
    Newcastle upon Tyne
    Posts
    2,978
    It depends.

    In your case it is not a disk to disk cloning as you "do" have other partitions in the original disk which is not available in the target (or recipient) disk. Thus the first sector or MBR of your two disks will be different.

    What I tend to do in such a case would be

    (A) Make a back up copies of the partitions that I do not want to clone into another source/disk

    (B) Print out a copy of the "fdisk -l" output for the source disk

    (C) Delete the unwanted partitions temporary. Using command like fdisk, cfdisk etc

    (D) Boot up the source disk to make sure everything is working as expected but do not touch the deleted partition area which should remain unallocated or blank.

    (E) Do a disk to disk cloning. Remove the source and boot up the target disk.

    (F) When both disks are working as expected boot up the source (without the target attached). Create the deleted partitions in exactly the same positions and sizes, using the previous "fdisk -l" as a guide.

    (G) Reboot and check everything back to normal in the source disk.

    The deletion of the partitions only alter the partition table and the data has not been destroyed (in fact not a single byte would be touched inside the partitions) if the unallocated space is not formatted. The partition table can be reinstated and all data will be recoverable. This is OK as I have done it loads of time.

    I advocate disk to disk cloning because no MS system would not refuse it. The integrity of the partition table is vital and as long as there is no change a MS Windows would accept itself being at a new disk with a different hardware serial number. A hard disk is liable to failure so M$ must allow its systems transferrable if the entire operating system is not changed which would be the case of a disk to disk cloning.
    Last edited by saikee; 03-25-2010 at 09:40 AM.
    Linux user started Jun 2004 - No. 361921
    Using a Linux live CD to clone XP
    To install Linux and keep Windows MBR untouched
    Adding extra Linux & Doing it in a lazy way
    A Grub menu booting 100+ systems & A "Howto" to install and boot 145 systems
    Just cloning tips Just booting tips A collection of booting tips

    Judge asked Linux "You are being charged murdering Windoze by stabbing its heart with a weapon, what was it?" Replied Linux "A Live CD"

  7. #142
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13

    Per your second nice reply...

    And I have decided on disk to clean disk on your good advice.
    And will not persist on this topic after this final inquiry .

    But -- sticking [hypothetical] wonderment Re:
    as you "do" have other partitions in the original disk which is not available in the target (or recipient) disk. Thus the first sector or MBR of your two disks will be different.
    My source/origin disk has NOTHING on it but the FAT32 D: utility (first)
    and the ntfs C: operating main (second).

    I proposed:
    To first fill a clean first FAT32 partition on the recipient/target disk with the D: contents from the source/origin; as I did before.
    To then fill a clean second ntfs partition on the recipient/target disk with the C: contents from the source/origin.
    The target/recipient disk DOES/would have some other stuff...but only on latter partitions.

    I am dependent on a live/CD Parted Magic for most of these machinations, having had good luck
    with that external operator and uncomfortable as generally I am with command line operation
    (absent specific, almost cut & paste guidelines), so....'nuff sed.

    And ALL of that said...I am done troubling this issue and you, who I am sure have much better things to do.
    I wish there were a way to donate to your nice resource but have been politely declined in that.
    So --cheers anyway, then and most gratefully, yours...

  8. #143
    Join Date
    Jun 2004
    Location
    Newcastle upon Tyne
    Posts
    2,978
    has_te,

    In your original proposal the target disk has already some partition inside and you are cloning another "c" and "d" drives onto it.

    The MS system when booted up in the target will find itself with a different partition table. I am not sure if Xp can tolerate it but Vista and Win7 will throw it out as these systems suspect a virus attack as the partition table has been compromised or has been changed without its knowledge.

    Your original scheme is also hard to execute because your source disk once formatted will have at least the first partition claiming the first primary position. The next partition added can only take up the second primary position. A bootloader finds its operating system by the hard disk address and so an operating system installed at the beginning of a hard disk, for being the first partition, will not boot if its position has been shifted downward in the hard disk because the boot loader will load the wrong information by going to the right original address of the hard disk.
    Linux user started Jun 2004 - No. 361921
    Using a Linux live CD to clone XP
    To install Linux and keep Windows MBR untouched
    Adding extra Linux & Doing it in a lazy way
    A Grub menu booting 100+ systems & A "Howto" to install and boot 145 systems
    Just cloning tips Just booting tips A collection of booting tips

    Judge asked Linux "You are being charged murdering Windoze by stabbing its heart with a weapon, what was it?" Replied Linux "A Live CD"

  9. #144
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13
    greetings [and I quote], ...Sir Grub Master-Guru
    Bit tardy getting back, slow spring here in Montana, like yours
    on the Island...I've heard, but outdoor stuff demanded Attn:.

    Very much appreciate your last post. Makes perfect sense, as always.
    'Bout 'given up anyway...just going to watch the health of the old/fullish disk
    and keep out the crap.

    But may get up the courage to have another go. Put the D: first and the C: second
    as I tentatively proposed. The Linux live CDs seem able to retrieve any boot screw up
    I do engineer (which I did in the last effort) and M$ does do a native boot...so got a get-out.
    And the amazing data dump facility, of course...and perhaps use "the default block size 512". (I'll let you know if I can get it working.)
    But that's for that nonce..Meantime thnx again for all the great instructionals and your
    current spate of patient with this latterly unwashed
    ---From Montana then, Happy Spring to the Isles.

  10. #145
    Join Date
    Apr 2010
    Posts
    3

    Confusion is Unavoidable

    First, Hi to all and a big thanks to everyone for submitting the results of their experiments, both successful and not! I see, referring to article #80 that caution is indicated when disk geometries are obviously different. The clue in that case was the number of heads in the old drive = 16, which probably means that it was originally intended to operate in the CHS environment.

    Newer drives set up for LBA operation will show 255 heads, so I believe that file descriptors from old to new just don't get you to the same place on the new drive. Perhaps using the methods outlined here would work - I haven't needed to try it out:

    http://www.nilbus.com/linux/disk-copy.php

    Mostly, the problems crop up with M$ systems as we already know. I also wanted to mention that while most people like to think of a new hard disk as a blank piece of paper, all drives are not created equal. When EIDE came about, all the bad track tables, secret hidden partition information with proprietary security programs now reside in the drive electronics and are not accessible to mortal beings. Manufacturers can put stuff anywhere they please and it will not be detectable because orders must go through the EIDE electronics before they get to the disk. I came across this weirdness while cloning $TEEVO drives, which must see certain data in certain places to operate.

    But fortunately, it looks as if 2 new LBA-style drives will accept a simple dd command successfully (look for 255 heads in the fdisk -l output). Thanks again to everyone for the very interesting discussion. It has made me very thoughtful about the whole subject, but especially appreciative for the existence of Linux.

    -Kavi

  11. #146
    Join Date
    Jun 2004
    Location
    Newcastle upon Tyne
    Posts
    2,978
    Kavi,

    Welcome to Justlinux!


    Your point of some bad track table and other manufacturer information could reside in the electrontic interface of the hard disk is credible and possibly done by many manufacturers. However such information is not accessible by "dd". What Linux can get is exactly what the hard disk manufacturer's declared number of cylinders, heads and sectors. dd and every operating system are not normally able to get into the on-board management information of the hard disk.

    If a target disk is of a different manufacturer to the source disk the electronic data can possibly be located physically at different locations between the two hard disks but operational wise they are in the exact numbers of CHS. Thus the data will have to be identical to the operating system, as long as the total number of CHS of the sourced data can be accommodated in the target disk.

    If the whole hard disk is cloned both the source and the target will have identical CHS numbers so it doesn't matter what the original arrangement is. In Linux the CHS can be freely altered by the users although I do not recommend messing around with it. If the hard disk is cloned it simply inherits what the source disk has got in the first place.

    M$ and other systems can use different geometries than Linux in the partition table and dead spaces can occur. These dead spaces will be the same and faithfully mirrored if the disk is cloned. Again dd will succeed.

    Currently I own over 70 hard disks in various format and sizes and have never failed even once in the cloning operation which to me is created for the benefit of the M$ systems. Linux permits migration of its systems and there is no need to clone Linux.

    Cloning the hard disk is convenient because the command is so simple that I have taught some teenagers to try it themselves. The difficult part is usually the physical swapping of the hard disks.
    Linux user started Jun 2004 - No. 361921
    Using a Linux live CD to clone XP
    To install Linux and keep Windows MBR untouched
    Adding extra Linux & Doing it in a lazy way
    A Grub menu booting 100+ systems & A "Howto" to install and boot 145 systems
    Just cloning tips Just booting tips A collection of booting tips

    Judge asked Linux "You are being charged murdering Windoze by stabbing its heart with a weapon, what was it?" Replied Linux "A Live CD"

  12. #147
    Join Date
    Apr 2010
    Posts
    3

    Thanks!

    Thanks for your comments. I have been using Acronis to make a backup of my XP disk for quite some time. The simpler dd method works perfectly also and XP does not complain or ask to be rebooted. (Maybe that problem starts with $Vista). My comments about all hd not being equal had to do with the fact that certain particular drives from certain manufacturers would not boot after a dd clone. I am not skillful enough to know why, but I suspected it must have been related to the need to position data in precise places. In any case, I am a bit wiser now thanks to all the great info in these threads.

    Kavi

  13. #148
    Join Date
    Jun 2004
    Location
    Newcastle upon Tyne
    Posts
    2,978
    Kavi ,

    If you have a hard disk that cannot be cloned using dd (this is to say it doesn't boot as the original) I would be interested to know.

    The nice thing about dd, which is available in every Unix and Linux system, is that it clones the hard disk sector by sector. As such the type of the operating system or the data in the hard disk is immaterial because it is the raw binary data that is being cloned. Theoretically it is impossible for a cloned copy to behave anything other than its original source.

    The dd operation is laughingly simple and is no more than a basic operation with a hard disk. M$ systems do not provide such a simple function and Corporations like Acronis and Norton Ghost make millions on the ignorance of the M$ system users.
    Linux user started Jun 2004 - No. 361921
    Using a Linux live CD to clone XP
    To install Linux and keep Windows MBR untouched
    Adding extra Linux & Doing it in a lazy way
    A Grub menu booting 100+ systems & A "Howto" to install and boot 145 systems
    Just cloning tips Just booting tips A collection of booting tips

    Judge asked Linux "You are being charged murdering Windoze by stabbing its heart with a weapon, what was it?" Replied Linux "A Live CD"

  14. #149
    Join Date
    Feb 2010
    Location
    Montana
    Posts
    13
    As a simple adendum,
    I did find it interesting in looking at the partition table that the old Samsung
    source disk in my recent foiled effort had (or claimed to have) 242 heads
    While what the much newer WD had is the apparently standard 244 heads.
    (Not that any of that had any impact on my own experience.)
    -But nice to see this thread is still so active-..
    cheers all & best wishes in any doing of the dd

  15. #150
    Join Date
    Apr 2010
    Posts
    3
    Hello,
    I'd be happy to supply you with some extra information regarding stubborn drives. But rather than filling the thread with distracting information I would like to PM you directly and let you decide on its value. Currently, the board will not allow me to send a PM, however.

    Kavi

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •