View previous topic :: View next topic |
Author |
Message |
SarahS93 l33t
Joined: 21 Nov 2013 Posts: 728
|
Posted: Sun Jun 09, 2024 1:21 am Post subject: split a file like a raid0 |
|
|
i put a videofile in a luks container with detached header. that is not enough for me.
what ways are there to split the luks file by put every second byte into file2 and all other not second bytes into file1?
like a (software) raid0 with 2 files/disks.
software raid0 i dont like becouse there are raid superblocks with them are visible that file1 and file2 are from a raid.
byte1 -> file1
byte2 -> file2
byte3 -> file1
byte4 -> file2
byte4 -> file1
....
or not bytes, chunks a little bit bigger?
i think security through obscurity is what i am looking for.
is there a way to add file1 and file2 on the fly together to mount them and use it, like a raid0 with mdadm? |
|
Back to top |
|
|
Banana Moderator
Joined: 21 May 2004 Posts: 1721 Location: Germany
|
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54577 Location: 56N 3W
|
Posted: Sun Jun 09, 2024 9:19 am Post subject: |
|
|
SarahS93,
There is no need to have raid superblocks. You need to have that data in a notebook though.
The very old raid format supported external superblocks or no superblqcks.
The drawback is that only manual assembly is possible and the raid elements must be listed in the right order at assemble time.
Raid0 gets you 'chunksize' pieces, not bytes. You can also have a lot more that two elements in a raid0
If you want to play, make a few empty files.
Use losetup to attach them to /dev/loopX
Tell mdadm to assemble the /dev/loop devices into a raid set. Do not use --create, as that writes the metadata that you don't want.
Make a filesystem on the raid device.
Raid5 will give you redundant data unless you use the special device missing for one raid element. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
SarahS93 l33t
Joined: 21 Nov 2013 Posts: 728
|
Posted: Fri Jun 14, 2024 3:26 pm Post subject: |
|
|
fallocate -l 10G datei1
fallocate -l 10G datei2
fallocate -l 10G datei3
losetup /dev/loop1 datei1
losetup /dev/loop2 datei2
losetup /dev/loop3 datei3
mdadm --assemble /dev/md0 --level=0 --raid-devices=3 /dev/loop1 /dev/loop2 /dev/loop3
mdadm: :option --level not valid in assemble mode
mdadm --assemble /dev/md0 /dev/loop1 /dev/loop2 /dev/loop3
mdadm: no recogniseable superblock on /dev/loop1
mdadm: /dev/loop1 has no superblock - assembly aborted
do i need "--build" first? and after that "--assemble"?
( https://raid.wiki.kernel.org/index.php/A_guide_to_mdadm#Build
mdadm --build /dev/md0 --level=0 --raid-devices=3 /dev/loop1 /dev/loop2 /dev/loop3
mdadm: RUN_ARRAY failed: Invalid argument
what do i make wrong?
:edit
grep "CONFIG_MD_RAID" /usr/src/linux/.config
Code: | # CONFIG_MD_RAID0 is not set
# CONFIG_MD_RAID1 is not set
# CONFIG_MD_RAID10 is not set
# CONFIG_MD_RAID456 is not set | is that the problem?! |
|
Back to top |
|
|
SarahS93 l33t
Joined: 21 Nov 2013 Posts: 728
|
Posted: Sat Jun 15, 2024 4:53 am Post subject: |
|
|
with set "CONFIG_MD_RAID0=y" it works
Code: | mdadm --build /dev/md0 --level=0 --raid-devices=3 /dev/loop1 /dev/loop2 /dev/loop3
mdadm: array /dev/md0 built and started. |
Code: | mdadm --detail /dev/md0
/dev/md0:
Version :
Creation Time : Sat Jun 15 06:54:48 2024
Raid Level : raid0
Array Size : 31457280 (30.00 GiB 32.21 GB)
Raid Devices : 3
Total Devices : 3
State : clean
Active Devices : 3
Working Devices : 3
Failed Devices : 0
Spare Devices : 0
Layout : -unknown-
Chunk Size : 64K
Consistency Policy : none
Number Major Minor RaidDevice State
0 7 1 0 active sync /dev/loop1
1 7 2 1 active sync /dev/loop2
2 7 3 2 active sync /dev/loop3 |
with
Code: | mdadm --verbose --assemble /dev/md0 /dev/loop1 /dev/loop2 /dev/loop3
mdadm: looking for devices for /dev/md0
mdadm: no recogniseable superblock on /dev/loop1
mdadm: /dev/loop1 has no superblock - assembly aborted |
it doesnt work
i do not understand "assemble"?
is "Chunk Size : 64K" OK?
is the build-way OK? |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54577 Location: 56N 3W
|
Posted: Sat Jun 15, 2024 11:12 am Post subject: |
|
|
SarahS93,
Quote: | is "Chunk Size : 64K" OK? |
The Chunk Size is the amount of data written to one device before the next device is used.
In your case, that's 64k each to dev/loop1 /dev/loop2 /dev/loop3 before it repeats ... dev/loop1 /dev/loop2 /dev/loop3 ...
64k is a good start. It wants to be an integer multiple of both the filesystem block size and the underlying block device black size.
Writing partial blocks or partial chunks is slow.
I've not used raid without super blocks. Just read about it.
Maybe you need to pass the metadata version too as that determines where the filesystem starts on the raid set.
-- edit --
man mdadm: | Build Build an array that doesn’t have per-device metadata (superblocks). For these sorts of arrays, mdadm cannot differ‐
entiate between initial creation and subsequent assembly of an array. It also cannot perform any checks that appro‐
priate components have been requested. Because of this, the Build mode should only be used together with a complete
understanding of what you are doing. |
You always build the array. Assemle checks that the components do form a bona fide array which requires the superblock data. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
|