On 2/18/22 4:04 PM, wicklowham wrote:
> On 18/02/2022 18:39, NP2 wrote:
>> Nic wrote:
>>> Is there a preferred way to copy a bootable flash drive to another
>>> flash drive?
>>>
>>> The FD I have was made using Ventoy and is bootable and has 9 iso's
>>> on it.
>>>
>>>
>>
>> On a related note: be careful with Mint's option "make bootable USB
>> stick", it doesn't always work. I recently tried to make a bootable
>> USB stick from a Windows 10 iso using this option, but the stick
>> ended up having one single udf file system instead of the expected
>> two partitions FAT16 + NTFS. Needless to say it didn't boot at all.
> ==============
> Sofar I have not had any problems with making a bootable flashdisk on
> Linux Mint . A right hand click on a downloaded .iso file starts the
> process ......but it does NOT result in  a persistant
> flashdisk.Recently I did it following downloading Zorin OS 16 now
> installed on a miniPC
>
> Frank   in County Wicklow -Ireland

You are talking about 'burnin' a drive, I was asking about reproducing a
drive.

Although I will add Zorin 16 to my list on my Ventoy FD.

Thank You Ireland.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

pinnerite wrote:
> I found that 16384 was optimal for cloning a 2TB drive
> status=progress displays statistics every 30 seconds (I think).

Interesting that 16KiB would be "ideal". Usually the curve peaks in the
MiB ranges (e.g. 4MiB for USB drives).

Also, to save typing (and math errors), you can use the suffixes
'K','M','G' (etc) for multiples of 1024 bytes (likewise KB,MB,GB(...)
for multiples of 1000 bytes).

-----BEGIN PGP SIGNATURE-----
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=tC2s
-----END PGP SIGNATURE-----

--
|_|O|_|
|_|_|O| Github: https://github.com/dpurgert
|O|O|O| PGP: DDAB 23FB 19FA 7D85 1CC1 E067 6D65 70E5 4CE7 2860

On 2/18/22 18:26, Nic wrote:
> On 2/18/22 12:10 PM, Dan Purgert wrote:
> Nic wrote:
>>>> Is there a preferred way to copy a bootable flash drive to another flash
>>>> drive?
>>>>
>>>> The FD I have was made using Ventoy and is bootable and has 9 iso's
>>>> on it.
>>>>
> I'd try dd first.
>
>
>>
> Is there a front end (gui) for dd yet?

I suppose that Etcher will do that.

>

--
“The further a society drift from the truth,
the more it will hate those who speak it”

George Orwell

On 2/19/2022 6:50 AM, Dan Purgert wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
>
> pinnerite wrote:
>> I found that 16384 was optimal for cloning a 2TB drive
>> status=progress displays statistics every 30 seconds (I think).
>
> Interesting that 16KiB would be "ideal". Usually the curve peaks in the
> MiB ranges (e.g. 4MiB for USB drives).
>
> Also, to save typing (and math errors), you can use the suffixes
> 'K','M','G' (etc) for multiples of 1024 bytes (likewise KB,MB,GB(...)
> for multiples of 1000 bytes).

Even 8KiB works on modern drives.

Use the drive size obtained from fdisk or gdisk, before you start.
Use the "factor" command, so conveniently provided on every Linux
distro, to work out possible values for bs= .

Here, I test whether I really understand what "dd" is doing.
I'm counting the bytes transferred from the source, to see
if it matches the fdisk disk size info.

sudo dd if=/dev/sda bs=a_bad_choice | wc -c # a check to see how
# many bytes are transferred
# from a source device

sudo dd if=/dev/sda bs=5632 | wc -c # did you get all the data ?
# What does (+1) mean when the
# transfer stats are printed out ?

On old (IDE ribbon cable era) drives, bs=221184 is a good value
for those.

You can get platter-speed-transfer on modern drives, with bs=8192.

You don't get platter speed transfer with bs=512, typically
less speed is the result.

On flash drives, you want transfer sizes larger than the page size.
There are about three page sizes for flash chips. The page size
affects the granularity of operations on flash chips ("OPs per second"),
which may account for the size not "zooming off to infinity and beyond".

If you use bs=512 on a flash drive, there might be unnecessary
write amplification as a result. Since the Linux caches are
working hard, even with flash, this might be inadvertently
taken care of, for you. You never know. We would need to use
a USB packet tracer, to see what is really going on. On OSes,
instrumentation for performance measure, is pretty patchy, and
lots of stuff is hidden from us.

The flash drive does not advertise its page size, nor does it
give the part number of the flash chip. Opaque, and scummy.

Paul