Oh... Oh well. Thanks a million for all your time, effort, and consideration, Awesome. I appreciate it.

Cbeers!

Thank you for saying so, Bangoram!! :-)

Hello Sir,

This has reference to the web page: "Learn The DD Command Revised"


> "Make an iso image of a CD"

Done every thing as under,

-------------------------------------------------------------------------------------------------------------------------------------------
knoppix@localhost:~$ su
Password:
root@localhost:/home/knoppix# dd if=/dev/cdrom of=mycd.iso bs=2048 conv=notrunc
358254+0 records in
358254+0 records out
733704192 bytes transferred in 354.825077 seconds (2067791 bytes/sec)
root@localhost:/home/knoppix#

mycd.iso has permissions-rw-r--r--)to chage the permissions:
root@localhost:/home/knoppix# chmod a+rwx mycd.iso
root@localhost:/home/knoppix#
Now mycd.iso has permissions-rwxrwxrwx)

to mount the image,
root@localhost:/home/knoppix#mkdir /mnt/mycd
root@localhost:/home/knoppix#

root@localhost:/home/knoppix#kedit /etc/fstab
added the line /home/knoppix/mycd.iso /mnt/mycd iso9660 rw, user, noauto 0 0 and saved it.

root@localhost:/home/knoppix# mount -o loop /mnt/mycd
root@localhost:/home/knoppix#

The file system IS available as files & directories.
---------------------------------------------------------------------------------------------------------------------------------------------

however sorry to inform you that,

can NOT edit "mycd.iso" as it gets mounted with "-r-xr-xr-x" permissions!

Prof S W Damle

Email : swdamle@sancharnet.in

You are correct. I don't believe the post states the image can be edited this way.

The images are not writable back when doing the things with mount point right.I believe it.But is there any other method that do the same Awesomemachine?

In practice,mounting the image somewhere,then edit the mount point likely to change the *image* also.is that possible?

There are ways to hack iso9660 images, and make modified CD's; for a distribution CD, or live CD; for instance. This type of hacking covers a broad range of possibilities and methods, but there are few common points in all. The contents of the CD image file must be copied from the loop device to an actual working file system. Changes must be made in ext3, reiserfs, etc.. Mkisofs is a powerful tool for creating CD image files, which can later be burned to a CD, out of a directory on a hard drive. One possible reason for editing an iso image file is to customize a linux distribution for certain hardware it won't currently install on. This involves changing /boot/initrd on the CD. There are other reasons, such as making a live CD, or customizing Knoppix. These subjects, and more, are covered in http://www.google.com . If you have a web browser, which you must if you are reading this, the answers you seek can be found on google. Maybe I'll make a post on hacking CD image files in the future. The dd tutorial is quite steady work, but if I see enough requests for CD image file information I'll eventually put something together and make a new thread.

Hi,

I am back on this after a break (got busy with other stuff. I still got the no more room on device error when I left out notrunc.


Before I go further, I want to clarify. Am I filling the free space with zeros on the hda drive (which is the drive that is currently running the machine) or the sda drive (the, new smaller drive)? Before or after I do the dd if=/dev/hda of=/dev/sdd.....?

>>dd if=/dev/zero > /home/sam/bigfile.file<<<

on this command, I am confused, how does it know which drive to do the /zero? meaning is if=/dev/zero shorthand for something like if=/dev/hda/zero or if=/dev/sdd/zero???

Thanks.
jon

Read back a few posts for the real story on notrunc. The device /dev/zero is the zero device. Whenever the system needs zeroes it gets them from the zero device, /dev/zero. There is also a random device, /dev/random, and a null device, /dev/null. Computers of all types use zeroes, nulls, and random data; and one of the things an operating system needs to do as part of its routine is provide these. Linux provides them with device files, making them manually accessible.

wonderful post mate, although i might never us the command to it's full extent, it is a really useful thing.

Glad you ran into it, kniwor

Hi.
Excelent post here.

I have copied a debian system runing in a vmware to a physical system using dd.
don't remember the command i typed... but it runed succesefully...
the VM disk was 16GB, the real one 40GB

it is now like this:
i only did dd on the hda1 not whole hda, cause i had already made the partitions...
if i run fdisk:

so partition table is correct, why df shows only 16gb?!

thanks

actually i will very much also like to know what has caused this, it will help me understand things a lot, and can someone point me to a tutorial telling what sectors of HDD is partitioning table/boot file etc...etc...etc.. in modern HDD's or how to query all this for a HDD, i read some tutorials but they seemed a bit old, for conventional old small sized (7.8G) HDD's. thanks in advance.

The way fdisk and df work is slightly different. Fdisk is a lower level utility, and more accurate. What I would do is this, in /etc, as root:

dd if=/dev/zero > /etc/bigfile.file

and fill the all free space with a big file. Dd will tell you when there is no space left on /dev/hd1. Then run df and fdisk, also run:

rm /etc/bigfile.file

to remove the big file. This method also works to overwrite all the free space on a partition, to avoid recovery of deleted files.

Although this is shortly to change, the partition table is in sector 1 on a HDD. The boot sector is sector 64, and data begins after the boot sector. There is metadata, which explains things about files, such as permissions, which is all grouped together in modern file systems. To get information about a HDD hardware you use:

hdparm

If you want to read the different parts of the disk you do something like:

dd if=/dev/sda count=1 | hexdump -C

Will show you sector 1, MBR, on the terminal screen. If you want to see the boot sector:

dd if=/dev/sda skip=63 | hexdump -C | less

will bring you to the beginning of the boot sector. You can use the space bar to scroll down. But, it isn't very interesting to most people to just look at. If you want to find valuable parts of the drive you can search for the magic signature, "55 AA" like this:

dd if=/dev/sda | hexdump -C | grep '55 AA'

The resulting output will give you a byte offset where each instance of "55 AA" is located. You then need to convert this offset to decimal from hexidecimal, divide it by 512, and use this number, minus 1, without fractional parts, as your skip= parameter.

Say you found "55 AA" at 4fb18000h. That's 5,632,000,000d, with any of the many calculators that come with linux. h and d stand for hexidecimal and decimal. So, divide 5,632,000,000 by 512 = 11,000,000. So:

dd if=/dev/sda skip=10999999 | hexdump -C | less

should take you to the sector where "55 AA" is located. You might find some interesting stuff, like the phantom MBR MS Windows puts at the end of the HDD, which never gets overwritten. You can find every partition and MBR by looking for "55 AA". Please note: 55 AA is hexidecimal. When passing hexidecimal to grep there must be a space between each byte: 55<space>AA, like that, but don't actually type it out or use less than and greater than signs. Use an actual space.

thanks for that tip...
i've made that but things remain the same

it seems that in the partition table it is marked as 37GB cause I partition that before dd, then i only made dd to the hda1, did you understand?

but didn't these 37535841 mean 37GB?

thanks for answering