Swap 32gb Cf For Microdrive, Easy Way to Do This... Options

[ArchiMark]

Decided to take the plunge and got a new 32GB A-Data CF Card today....



Want to do this swap the right way and hopefully easy way....

I assume that I can 'clone' my current Debian EABI setup on my existing 6GB microdrive over to the new 32GB CF card, right?

Would this be the right way to do it?

1) Format new 32GB card and make 2 partitions; ext3 (/dev/hdc1 /mnt/cf) and a small swap (512MB on /dev/hdc2 /mnt/cf2).

2) Copy over current microdrive debian contents on hda2 over to new ext3 partition on 32GB CF card by doing:

CODE

#dd if=/dev/hda2 of=/dev/cf

* Note: hda1 on my existing microdrive is my swap partition.

So, is this correct way???

Thanks!

Mark

This post has been edited by ArchiMark: Apr 1 2008, 07:24 PM

[speculatrix]

Decided to take the plunge and got a new 32GB A-Data CF Card today....

no, you definitely don't want to use "dd" to copy a file system between different types of media unless you guarantee the number of blocks in the partition is the same or larger on the target.

anyway, "cat" is faster - it uses more buffering!

the best way to copy a filesys is like this:

cd /mnt/partition1
tar cf - | (cd /mnt/partition2 ; tar xf - )

tar will copy all files and devices, preserve the datestamps and protections etc.

if you were l33t, you could probably also do it with cpio but I'm too lazy to look it up.

so, just partition the CF appropriately. oh, do NOT use ext3, the journalling will kill the card faster.

[ArchiMark]

anyway, "cat" is faster - it uses more buffering!

I just realized that I don't see 'cat' below in your instructions???

Also, I thought that 'cat' is for just viewing the content of a file?

the best way to copy a filesys is like this:

CODE

cd /mnt/partition1
tar cf - | (cd /mnt/partition2; tar xf - )

tar will copy all files and devices, preserve the datestamps and protections etc.

So, just want to make sure I understand this correctly for my situation of copying my hda2 partition with Debian EABI (hda1 is my swap partition) on it over to new partition on 32GB CF card....

So, here's the existing and new setups:

Existing 6GB Microdrive
hda1 is Swap Partition
hda2 is Debian EABI Partition

New 32GB CF Card
cf1 will be new Debian EABI Partition
cf2 will be new Swap Partition

CODE

cd /mnt/cf1
tar /mnt/hda2 - | (cd /mnt/cf1; tar xf - )

Does this look right?

Thanks!

[speculatrix]

anyway, "cat" is faster - it uses more buffering!

I just realized that I don't see 'cat' below in your instructions???

Also, I thought that 'cat' is for just viewing the content of a file?

no, you could do, for example, "cat /dev/hda > /dev/hdb" to clone an entire hard drive

So, here's the existing and new setups:

Existing 6GB Microdrive
hda1 is Swap Partition
hda2 is Debian EABI Partition

New 32GB CF Card
cf1 will be new Debian EABI Partition
cf2 will be new Swap Partition

CODE

cd /mnt/cf1
tar /mnt/hda2 - | (cd /mnt/cf1; tar xf - )

Does this look right?

Thanks!

you're missing a dot

cd /mnt/hda2
tar cf - . | (cd /mnt/cf1 ; tar xf - )


you cd into the source directory so that when you do the tar the current directory is the "root" of all the files; the "-" in the tar command outputs the stream to stdout; the stream is piped into a separate shell (the brackets) which starts in the destionation directory and unpacks the incoming stream from stdin.

the same trick works over an ssh session.

[ArchiMark]

anyway, "cat" is faster - it uses more buffering!

I just realized that I don't see 'cat' below in your instructions???

Also, I thought that 'cat' is for just viewing the content of a file?

no, you could do, for example, "cat /dev/hda > /dev/hdb" to clone an entire hard drive

So, here's the existing and new setups:

Existing 6GB Microdrive
hda1 is Swap Partition
hda2 is Debian EABI Partition

New 32GB CF Card
cf1 will be new Debian EABI Partition
cf2 will be new Swap Partition

CODE

cd /mnt/cf1
tar /mnt/hda2 - | (cd /mnt/cf1; tar xf - )

Does this look right?

Thanks!

you're missing a dot

cd /mnt/hda2
tar cf - . | (cd /mnt/cf1 ; tar xf - )


you cd into the source directory so that when you do the tar the current directory is the "root" of all the files; the "-" in the tar command outputs the stream to stdout; the stream is piped into a separate shell (the brackets) which starts in the destionation directory and unpacks the incoming stream from stdin.

the same trick works over an ssh session.

THANKS for the explanation/clarification, speculatrix!!

Think I got it now....