4

I am trying to encrypt an external hard drive on Ubuntu 14.04 followig this guide. I started by formatting to ext4:

$ sudo mkfs.ext4 /dev/sde1
mke2fs 1.42.9 (4-Feb-2014)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
9773056 inodes, 39072470 blocks
1953623 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=4294967296
1193 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks: 
    32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 
    4096000, 7962624, 11239424, 20480000, 23887872

Allocating group tables: done                            
Writing inode tables: done                            
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done     

And then proceeded with the initialisation, but it returns this error:

$ sudo cryptsetup -y -v luksFormat /dev/sde1

WARNING!
========
This will overwrite data on /dev/sde1 irrevocably.

Are you sure? (Type uppercase yes): YES
Enter passphrase: 
Verify passphrase: 
Cannot wipe header on device /dev/sde1.
Command failed with code 5: Cannot wipe header on device /dev/sde1.

So far I could not anything useful on the web regarding this error. Any ideas on what may be wrong?

Update I: answering the questions posed by Xen2050. I ran bablocks in write mode prior to formatting and no errors were reported.

I tried the ecryption again, paying more attention to systems messages. Here is the dmesg output right after connecting the drive:

$ dmesg
[ 3208.032228] usb 2-1.4: new high-speed USB device number 7 using ehci-pci
[ 3208.140990] usb 2-1.4: New USB device found, idVendor=059f, idProduct=0651
[ 3208.141001] usb 2-1.4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 3208.141024] usb 2-1.4: Product: LaCie Hard Drive USB
[ 3208.141031] usb 2-1.4: Manufacturer: LaCie
[ 3208.141037] usb 2-1.4: SerialNumber: 10000E000BD8A671
[ 3208.177576] usb-storage 2-1.4:1.0: USB Mass Storage device detected
[ 3208.178112] scsi4 : usb-storage 2-1.4:1.0
[ 3208.178183] usbcore: registered new interface driver usb-storage
[ 3209.176917] scsi 4:0:0:0: Direct-Access     SEAGATE  ST3160812A       3.AA PQ: 0 ANSI: 2
[ 3209.177561] sd 4:0:0:0: Attached scsi generic sg2 type 0
[ 3209.181342] sd 4:0:0:0: [sdb] 312581808 512-byte logical blocks: (160 GB/149 GiB)
[ 3209.182337] sd 4:0:0:0: [sdb] Write Protect is off
[ 3209.182348] sd 4:0:0:0: [sdb] Mode Sense: 53 00 00 08
[ 3209.183339] sd 4:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[ 3209.201618]  sdb: sdb1
[ 3209.229465] sd 4:0:0:0: [sdb] Attached SCSI disk

Then verified it is not mounted:

$ findmnt /dev/sdb
$ findmnt /dev/sdb1
$

Another try at initialisation:

$ sudo cryptsetup -y -v luksFormat /dev/sdb1

WARNING!
========
This will overwrite data on /dev/sdb1 irrevocably.

Are you sure? (Type uppercase yes): YES
Enter passphrase: 
Verify passphrase: 
Cannot wipe header on device /dev/sdb1.
Command failed with code 5: Cannot wipe header on device /dev/sdb1.

Two new lines show up in the log:

$ tail /var/log/syslog
Dec  8 09:18:20 MekanikDestruktiwKommandoh kernel: [ 3698.016311] end_request: critical target error, dev sdb, sector 0
Dec  8 09:18:28 MekanikDestruktiwKommandoh wpa_supplicant[1188]: wlan0: CTRL-EVENT-SCAN-STARTED 

Something seems going wrong with sector 0, but not at all clear what.

Update II: Trying the new suggestions by Xen2050. First writing zeros to sector 0:

$ sudo dd if=/dev/zero of=/dev/sdb1 bs=1M count=10
10+0 records in
10+0 records out
10485760 bytes (10 MB) copied, 0.564427 s, 18.6 MB/s

After this the cryptsetup initialisation still fails. wipefs returns a strange warning:

$ sudo wipefs -a /dev/sdb
wipefs: WARNING: /dev/sdb: appears to contain 'dos' partition table

And apparently does nothing. Then ran cryptsetup with the debug flag; it printed out a bunch of new stuff, but does not give more information regarding the error:

$ sudo cryptsetup --debug -y -v luksFormat /dev/sdb
# cryptsetup 1.6.1 processing "cryptsetup --debug -y -v luksFormat /dev/sdb"
# Running command luksFormat.
# Locking memory.
# Installing SIGINT/SIGTERM handler.
# Unblocking interruption on signal.

WARNING!
========
This will overwrite data on /dev/sdb irrevocably.

Are you sure? (Type uppercase yes): YES
# Allocating crypt device /dev/sdb context.
# Trying to open and read device /dev/sdb.
# Initialising device-mapper backend library.
# Timeout set to 0 miliseconds.
# Iteration time set to 1000 miliseconds.
# Interactive passphrase entry requested.
Enter passphrase: 
Verify passphrase: 
# Formatting device /dev/sdb as type LUKS1.
# Crypto backend (gcrypt 1.5.3) initialized.
# Topology: IO (512/0), offset = 0; Required alignment is 1048576 bytes.
# Generating LUKS header version 1 using hash sha1, aes, xts-plain64, MK 32 bytes
# Crypto backend (gcrypt 1.5.3) initialized.
# KDF pbkdf2, hash sha1: 356173 iterations per second.
# Data offset 4096, UUID 5fa9c58f-b047-4c9e-a6e8-26a9a433a438, digest iterations 43375
Cannot wipe header on device /dev/sdb.
# Releasing crypt device /dev/sdb context.
# Releasing device-mapper backend.
# Unlocking memory.
Command failed with code 5: Cannot wipe header on device /dev/sdb.
2
  • That's an onion in the ointment... any more descriptive messages in /var/log/syslog? /dev/sde1 isn't mounted while trying to luksFormat it? Maybe cryptsetup doesn't want to overwrite a mounted in-use partition... And it's still a "healthy" writeable drive? And the format to ext4 step should be after the luksFormat, but I think cryptsetup should be able to overwrite anything anyway, so that shouldn't stop it...
    – Xen2050
    Dec 6, 2015 at 16:48
  • How strange, badblocks writes ok, mkfs.ext4 writes ok, but cryptsetup fails trying to write to sector 0... maybe a cryptsetup bug, I wonder if sdb1 contains a sector 0, or if it's trying to write to the very start of sdb (actually I'm not sure if drives start at sector 1 or 0, writing to a non-existing sector sounds like a critical error...) (No access to a regular browser now or I'd google that sector 0 error + cryptsetup)
    – Xen2050
    Dec 8, 2015 at 11:03

3 Answers 3

4

I'm aware that this question is very old, but I was looking for a solution for the same error message, "Cannot wipe header on device", and found a solution that is not listed here: if the cryptsetup luksFormat output also includes the following:

WARNING: Data offset is outside of currently available data device.

then the partition is too small.

You're most likely to run into this if you are creating a partition that is meant to just contain a key file to encrypt other partitions.

It appears that with cryptsetup 2.1.0, the LUKS header takes up just under 16MiB (see below for details), so the partition size must 16MiB + the size of the key you want to store in it (3MiB, say).

For illustration, with an 8MiB partition, I saw the following error:

$ sudo cryptsetup luksFormat ${/dev/keypartition} --debug
[...]
# Formatting LUKS2 with JSON metadata area 12288 bytes and keyslots area 16744448 bytes.
[...]
WARNING: Data offset is outside of currently available data device.
Device wipe error, offset 8388608.
Cannot wipe header on device /dev/sda2.

Note that 12288 bytes + 16744448 bytes = 15.98 MiB.

With the size of /dev/sda2 increased to 19MiB, I successfully set it up as a LUKS partition.

Note: cryptsetup luksFormat also worked with dev/sda2 at 16MiB, but a subsequent cryptsetup luksOpen failed with

Requested offset is beyond real size of device /dev/sda2.

Increasing /dev/sda2 further to 32MiB fixed this issue, but when I tried to use the partition as a key file, cryptsetup luksFormat --key-file=/dev/mapper/<mountpoint of /dev/sda2> <device to encrypt> failed with

Maximum keyfile size exceeded.

There are two solutions to this: (1) set the size of the partition containing to key to 16MiB + key file size ensuring that the key file size is less than the maximum; (2) use the --keyfile-size option so cryptsetup luksFormat only uses some part of the key file.

3

One fellow solved this (or very similar) error by manually writing zeros to the first few megs of the target partition with a line similar to this:

 dd if=/dev/zero of=[target] bs=1M count=2

Or if it's just an overwriting FS problem, then wipefs might work instead of / with the above.

Also, another used the "--debug" option with cryptsetup to get more information, it's worth a try first.


The "Update II" info tests use /dev/sdb1 and then /dev/sdb, not sure it would make a difference though.

Another guy here says "I've had floppies, thumb drives and hard drives have 'sector 0' issues that were all fixed by going into a Linux Live CD and wiping the first hundred sectors or so with dd if=/dev/zero of=/dev/hdN bs=1024 count=1024 "

This bug report comment about another external USB hard drive and a "end_request: critical target error, dev sdb, sector 0" seems to say Ubuntu 12.04 worked while 14.04 didn't. Others say it's probably a bad or inadequate USB power supply or failing hardware (enclosure, wires, hard drive)

So mixed results & advice. I've not had the best luck with USB enclosures either, I'd be inclined to try a different enclosure or a direct to desktop/esata cable connection.

1

I was getting a very similar error message as recently as 2022 with 20.04.4 LTS when trying to cryptsetup a partition on a ~1TB external SSD:

Device wipe error, offset 0.
Cannot wipe header on device /dev/sda2.

Bizarrely enough, simply unplugging a bunch of USB devices (everything except keyboard / mouse / the external SSD) made the error go away.

1
  • 2
    Insufficient USB power seems to do this. I had a drive go offline under write load (2.5" HDD) on a liberated Chromebook. Too bad cryptsetup doesn't report that the lower layer disappeared - the error message becomes confusing requiring a dmesg scan to figure it out. Jul 27, 2022 at 18:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .