Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #245987 > unrolled thread

Conviently having netinst.iso on a 32GB flash drive

Back to article view | Back to linux.debian.user

Contents

  Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 11:10 +0100
    Re: Conviently having netinst.iso on a 32GB flash drive Joe <joe@jretrading.com> - 2022-03-06 12:00 +0100
      Re: Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 12:40 +0100
    Re: Conviently having netinst.iso on a 32GB flash drive songbird <songbird@anthive.com> - 2022-03-06 13:10 +0100
      Re: Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 13:50 +0100
        Re: Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 21:30 +0100
      Re: Conviently having netinst.iso on a 32GB flash drive "Thomas Schmitt" <scdbackup@gmx.net> - 2022-03-06 14:20 +0100
        Re: Conviently having netinst.iso on a 32GB flash drive "Thomas Schmitt" <scdbackup@gmx.net> - 2022-03-06 14:20 +0100
        Re: Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 15:50 +0100
          Re: Conviently having netinst.iso on a 32GB flash drive "Thomas Schmitt" <scdbackup@gmx.net> - 2022-03-06 17:30 +0100
            Re: Conviently having netinst.iso on a 32GB flash drive Richard Owlett <rowlett@cloud85.net> - 2022-03-06 21:30 +0100
        Re: Conviently having netinst.iso on a 32GB flash drive songbird <songbird@anthive.com> - 2022-03-06 21:00 +0100
        Re: Conviently having netinst.iso on a 32GB flash drive songbird <songbird@anthive.com> - 2022-03-06 21:00 +0100

#245987 — Conviently having netinst.iso on a 32GB flash drive

Currently I use
> dd if=netinst.iso bs=64M of=/dev/sdb

I would like way to copy it such that:
1. a legacy BIOS could launch it
2. Gparted would not complain about block size
3. there would be at least two partitions usable misc files

TIA

[toc] | [next] | [standalone]

#245988

On Sun, 6 Mar 2022 04:06:23 -0600
Richard Owlett <rowlett@cloud85.net> wrote:

> Currently I use
> > dd if=netinst.iso bs=64M of=/dev/sdb  
> 
> I would like way to copy it such that:
> 1. a legacy BIOS could launch it
> 2. Gparted would not complain about block size
> 3. there would be at least two partitions usable misc files
> 

This probably doesn't quite meet all your needs, but it's fairly new
and not widely known yet, you may want to investigate it anyway:

https://www.ventoy.net/en/index.html

It's new, so here may be dragons, but I've used it a bit and nothing
has broken yet. I have four Debian netinst.iso images (including an
i386) and winpe.iso on an 8GB microSD. Installation wipes the card, then
makes two partitions (one EFI), but the exfat iso partition is still
writable for files and directories. I don't know about the prospects of
shrinking the exfat partition, I don't use the format myself, but the
documents suggest you can reformat it to practically anything.

-- 
Joe

[toc] | [prev] | [next] | [standalone]

#245990

On 03/06/2022 04:52 AM, Joe wrote:
> On Sun, 6 Mar 2022 04:06:23 -0600
> Richard Owlett <rowlett@cloud85.net> wrote:
> 
>> Currently I use
>>> dd if=netinst.iso bs=64M of=/dev/sdb
>>
>> I would like way to copy it such that:
>> 1. a legacy BIOS could launch it
>> 2. Gparted would not complain about block size
>> 3. there would be at least two partitions usable misc files
>>
> 
> This probably doesn't quite meet all your needs, but it's fairly new
> and not widely known yet, you may want to investigate it anyway:
> 
> https://www.ventoy.net/en/index.html

Browsed the site. Looks very promising. It may not fulfill my 3rd point 
literally. But seems to have a solution for the problem I wanted to address.

Its ability to handle multiple ISO files may allow to solve a problem 
I've dreamed of attacking for a VERY long time.


> 
> It's new, so here may be dragons,

Dragons are where the fun is ;}
Thank you.

> but I've used it a bit and nothing
> has broken yet. I have four Debian netinst.iso images (including an
> i386) and winpe.iso on an 8GB microSD. Installation wipes the card, then
> makes two partitions (one EFI), but the exfat iso partition is still
> writable for files and directories. I don't know about the prospects of
> shrinking the exfat partition, I don't use the format myself, but the
> documents suggest you can reformat it to practically anything.
> 

[toc] | [prev] | [next] | [standalone]

#245992

Richard Owlett wrote:
> Currently I use
>> dd if=netinst.iso bs=64M of=/dev/sdb
>
> I would like way to copy it such that:
> 1. a legacy BIOS could launch it
> 2. Gparted would not complain about block size
> 3. there would be at least two partitions usable misc files


  one of the nice people here posted a program for doing this 
called make_isombr_part (or something similar).  see if you
can find it.  archive search for this group Feb-Mar 2017.

  i have two versions of it now, but it was a five years ago and
i don't recall if i made the changes or got a new version from
the author.  :)

  my somewhat incomplete notes look like:

=====
This program makes the rest of a USB stick available
to create another partition after you put an ISO image
on a USB stick.

the whole thing goes like:

  1. download netinst, or some other ISO image

  2. cp it to the USB stick (to the whole device)

   example:  (check for right device before clobbering)

#  cp /home/me/IWDownloads/debian-stretch-DI-rc2-i386-netinst.iso /dev/sd<X>


  3. add boot loader (syslinux)


  4.  make isombr partition


usage: ./make_isombr_part /dev/sd<X> auto 83 dry_run

remove dry_run when you actually want the changes to be made.

  5. create file system on what remains

# mkfs.ext4 -v -b 2048 -L FD02 /dev/sd<x><partition_number>

=====


  songbird

[toc] | [prev] | [next] | [standalone]

#245994

On 03/06/2022 05:43 AM, songbird wrote:
> Richard Owlett wrote:
>> Currently I use
>>> dd if=netinst.iso bs=64M of=/dev/sdb
>>
>> I would like way to copy it such that:
>> 1. a legacy BIOS could launch it
>> 2. Gparted would not complain about block size
>> 3. there would be at least two partitions usable misc files
> 
> 
>    one of the nice people here posted a program for doing this
> called make_isombr_part (or something similar).  see if you
> can find it.  archive search for this group Feb-Mar 2017.
> [SNIP]

First Duckduckgo hit is
https://lists.debian.org/debian-user/2017/03/msg01268.html

Post by Thomas Schmitt.

As tonight's weather is scheduled to be flooding rain followed by 
unspecified amount of snow I'll have time to chase down any updates to 
it and do initial experiments with the ventoy package suggested by Joe.

Thanks

[toc] | [prev] | [next] | [standalone]

#246009

On 03/06/2022 06:44 AM, Richard Owlett wrote:
> On 03/06/2022 05:43 AM, songbird wrote:
>> Richard Owlett wrote:
>>> Currently I use
>>>> dd if=netinst.iso bs=64M of=/dev/sdb
>>>
>>> I would like way to copy it such that:
>>> 1. a legacy BIOS could launch it
>>> 2. Gparted would not complain about block size
>>> 3. there would be at least two partitions usable misc files
>>
>>
>>    one of the nice people here posted a program for doing this
>> called make_isombr_part (or something similar).  see if you
>> can find it.  archive search for this group Feb-Mar 2017.
>> [SNIP]
> 
> First Duckduckgo hit is
> https://lists.debian.org/debian-user/2017/03/msg01268.html
> 
> Post by Thomas Schmitt.
> 
> As tonight's weather is scheduled to be flooding rain followed by 
> unspecified amount of snow I'll have time to chase down any updates to 
> it and do initial experiments with the ventoy package suggested by Joe.

Definite snow has been replaced by possible tornadoes.
The description of ventoy is beginning to intrigue me.
I'll pursue that venue.

[toc] | [prev] | [next] | [standalone]

#245995

Hi,

Richard Owlett wrote:
> > Currently I use
> > > dd if=netinst.iso bs=64M of=/dev/sdb
> > I would like way to copy it such that:
> > 1. a legacy BIOS could launch it

This should work fine after above dd run.


> > 2. Gparted would not complain about block size

Usually the partition editors complain about the weird mix of MBR partition
table, GPT, and Apple Partition Map (APM), of which the former two serve as
boot lures for EFI firmware and the latter is useless.

What complaints exactly do you get from gparted when trying to add a
new partition ?


> > 3. there would be at least two partitions usable misc files

songbird wrote:
> one of the nice people here posted a program for doing this
> called make_isombr_part (or something similar).

That was me.

Presentation of first version:
  https://lists.debian.org/debian-user/2017/03/msg01215.html

Richard Owlett's encounter first failed because the binary was amd64 and
his architecture was i386:
  https://lists.debian.org/debian-user/2017/03/msg01218.html
I then proposed to build it from source, which seems to have succeeded:
  https://lists.debian.org/debian-user/2017/03/msg01225.html


> i have two versions of it now, but it was a five years ago and
> i don't recall if i made the changes or got a new version from
> the author.  :)

Your request for a dry-run mode was fulfilled by a new version:
  https://lists.debian.org/debian-user/2017/03/msg01268.html
which needed a fix because of a misleading program message:
  https://lists.debian.org/debian-user/2017/03/msg01270.html
That's the version which is still available as
  http://scdbackup.webframe.org/make_isombr_part.c
  MD5 dd3e1a16e9593f908a1ce9ec848fd929

(My local version is slightly younger and seems to have been augmented
for being able to determine the USB stick size on FreeBSD, too. Dunno
whether this was for real use or just some finger exercises for myself.)


>   3. add boot loader (syslinux)

Why this step ?
The ISO brings a SYSLINUX/ISOLINUX boot loader for legacy BIOS (and GRUB
for EFI). The MBR of the ISO hops onto the ISOLINUX boot image which then
brings up the SYSLINUX boot menu.

What negative effect did you see when not performing this step 3 ?


Have a nice day :)

Thomas

[toc] | [prev] | [next] | [standalone]

#245996

Hi,

as usual, a little mistake:

I worte:
> That's the version which is still available as
>   http://scdbackup.webframe.org/make_isombr_part.c
>   MD5 dd3e1a16e9593f908a1ce9ec848fd929

That is the MD5 of the amd64 executable binary.
The MD5 of the source file is 34aa900801f65955a61cebf0280eeb3b .


Have a nice day :)

Thomas

[toc] | [prev] | [next] | [standalone]

#245997

On 03/06/2022 07:09 AM, Thomas Schmitt wrote:
> Hi,
> 
> Richard Owlett wrote:
>>> Currently I use
>>>> dd if=netinst.iso bs=64M of=/dev/sdb
>>> I would like way to copy it such that:
>>> 1. a legacy BIOS could launch it
> 
> This should work fine after above dd run.

I does.

> 
> 
>>> 2. Gparted would not complain about block size
> 
> Usually the partition editors complain about the weird mix of MBR partition
> table, GPT, and Apple Partition Map (APM), of which the former two serve as
> boot lures for EFI firmware and the latter is useless.
> 
> What complaints exactly do you get from gparted when trying to add a
> new partition ?

The error message block is titled
> It is not possible to create more than 1 primary partition

Gparted reports existing partition information to be:

Partition   | File system | Label               |  Size      | Used 
/dev/sdc1   | unknown     |Debian 11.2.0 amd64  |   4.00 KiB |       ---
unallocated | unallocated |                     |   1.98 MiB |       ---
/dev/sdc2   | fat16       |                     |   2.53 MiB |  2.52 MiB
unallocated | unallocated |                     |  29.81 GiB |       ---

> 
> 
>>> 3. there would be at least two partitions usable misc files
> 
> songbird wrote:
>> one of the nice people here posted a program for doing this
>> called make_isombr_part (or something similar).
> 
> That was me.
> 
> Presentation of first version:
>    https://lists.debian.org/debian-user/2017/03/msg01215.html
> 
> Richard Owlett's encounter first failed because the binary was amd64 and
> his architecture was i386:
>    https://lists.debian.org/debian-user/2017/03/msg01218.html
> I then proposed to build it from source, which seems to have succeeded:
>    https://lists.debian.org/debian-user/2017/03/msg01225.html

After successful compilation I don't recall having actually attempted to 
use it.

R. Owlett

> 
> 
>> i have two versions of it now, but it was a five years ago and
>> i don't recall if i made the changes or got a new version from
>> the author.  :)
> 
> Your request for a dry-run mode was fulfilled by a new version:
>    https://lists.debian.org/debian-user/2017/03/msg01268.html
> which needed a fix because of a misleading program message:
>    https://lists.debian.org/debian-user/2017/03/msg01270.html
> That's the version which is still available as
>    http://scdbackup.webframe.org/make_isombr_part.c
>    MD5 dd3e1a16e9593f908a1ce9ec848fd929
> 
> (My local version is slightly younger and seems to have been augmented
> for being able to determine the USB stick size on FreeBSD, too. Dunno
> whether this was for real use or just some finger exercises for myself.)
> 
> 
>>    3. add boot loader (syslinux)
> 
> Why this step ?
> The ISO brings a SYSLINUX/ISOLINUX boot loader for legacy BIOS (and GRUB
> for EFI). The MBR of the ISO hops onto the ISOLINUX boot image which then
> brings up the SYSLINUX boot menu.
> 
> What negative effect did you see when not performing this step 3 ?
> 
> 
> Have a nice day :)
> 
> Thomas
> 
> 
> 

[toc] | [prev] | [next] | [standalone]

#246004

Hi,

Richard Owlett wrote:
> The error message block is titled
> > It is not possible to create more than 1 primary partition

There should still 2 MBR partition table slots free. Both could become
primary partitions, although a provident partition editor could force
you to use the last partition as extended partition where more logical
partitions can be created.


> Gparted reports existing partition information to be:
>
> Partition   | File system | Label               |  Size      | Used
> /dev/sdc1   | unknown     |Debian 11.2.0 amd64  |   4.00 KiB |       ---
> unallocated | unallocated |                     |   1.98 MiB |       ---
> /dev/sdc2   | fat16       |                     |   2.53 MiB |  2.52 MiB
> unallocated | unallocated |                     |  29.81 GiB |       ---

That's quite some misperception.

  /sbin/fdisk -l debian-11.0.0-amd64-netinst.iso

says

  Device                           Boot Start    End Sectors  Size Id Type
  debian-11.0.0-amd64-netinst.iso1 *        0 772095  772096  377M  0 Empty
  debian-11.0.0-amd64-netinst.iso2       4060   9243    5184  2.5M ef EFI (FAT-12/

I guess that gparted's confusion is because of the nested partitions.
Another cause could be the GPT and APM debris. (It is the main job of
make_isombr_part to remove this.)

In order to check the theory about nested partitions you could delete
partition 2 which you don't need for BIOS booting.
E.g. by this interactive run:

  $ sudo fdisk /dev/sdb
  ... welcome messages , maybe use p to get the table shown ...

  Command (m for help): d
  Partition number (1,2, default 2): 2

  Partition 2 has been deleted.

  Command (m for help): w
  The partition table has been altered.
  Syncing disks.

Maybe gparted likes it better afterwards.


> >    http://scdbackup.webframe.org/make_isombr_part.c

> After successful compilation I don't recall having actually attempted to use
> it.

But
  https://lists.debian.org/debian-user/2017/03/msg01225.html
looks like you ran it and then inspected the resulting partition table:

>...> Device     Boot   Start      End  Sectors  Size Id Type
>...> /dev/sdb1  *          0  1140735  1140736  557M  0 Empty
>...> /dev/sdb2          4156     5531     1376  688K ef EFI (FAT-12/16/32)
>...> /dev/sdb3       1140736 15130623 13989888  6.7G 83 Linux


Have a nice day :)

Thomas

[toc] | [prev] | [next] | [standalone]

#246008

On 03/06/2022 10:25 AM, Thomas Schmitt wrote:
> [ *SNIP* ]
> I guess that gparted's confusion is because of the nested partitions.
> Another cause could be the GPT and APM debris. (It is the main job of
> make_isombr_part to remove this.)
> 
> In order to check the theory about nested partitions you could delete
> partition 2 which you don't need for BIOS booting.
> E.g. by this interactive run:
> 
>    $ sudo fdisk /dev/sdb
>    ... welcome messages , maybe use p to get the table shown ...
> 
>    Command (m for help): d
>    Partition number (1,2, default 2): 2
> 
>    Partition 2 has been deleted.
> 
>    Command (m for help): w
>    The partition table has been altered.
>    Syncing disks.
> 
> Maybe gparted likes it better afterwards.

*NO*    !
It does nothing *useful* for gparted.
The *_ONLY_* change is that the file type of /dev/sdc2 is now "unknown".

[toc] | [prev] | [next] | [standalone]

#246005

Thomas Schmitt wrote:
> Hi,

  :)


...
> songbird wrote:
>> one of the nice people here posted a program for doing this
>> called make_isombr_part (or something similar).
>
> That was me.
>
> Presentation of first version:
>   https://lists.debian.org/debian-user/2017/03/msg01215.html
>
> Richard Owlett's encounter first failed because the binary was amd64 and
> his architecture was i386:
>   https://lists.debian.org/debian-user/2017/03/msg01218.html
> I then proposed to build it from source, which seems to have succeeded:
>   https://lists.debian.org/debian-user/2017/03/msg01225.html
>
>
>> i have two versions of it now, but it was a five years ago and
>> i don't recall if i made the changes or got a new version from
>> the author.  :)
>
> Your request for a dry-run mode was fulfilled by a new version:
>   https://lists.debian.org/debian-user/2017/03/msg01268.html
> which needed a fix because of a misleading program message:
>   https://lists.debian.org/debian-user/2017/03/msg01270.html
> That's the version which is still available as
>   http://scdbackup.webframe.org/make_isombr_part.c
>   MD5 dd3e1a16e9593f908a1ce9ec848fd929

  :)  i have no recall of that any more (my brain's been
reused by many things over the past five years), but thank
you!  :)


> (My local version is slightly younger and seems to have been augmented
> for being able to determine the USB stick size on FreeBSD, too. Dunno
> whether this was for real use or just some finger exercises for myself.)
>
>>   3. add boot loader (syslinux)
>
> Why this step ?
> The ISO brings a SYSLINUX/ISOLINUX boot loader for legacy BIOS (and GRUB
> for EFI). The MBR of the ISO hops onto the ISOLINUX boot image which then
> brings up the SYSLINUX boot menu.
>
> What negative effect did you see when not performing this step 3 ?

  i have no idea or recall.  it may have been some fantasy on
my part.  my rough notes were written to make sure i had some
way of knowing what i was up to but i only did it the one time
back then and didn't need to use those notes again (or i'd 
have fixed them up).  for me it helps to have at least a stick
poking in about the right direction instead of having to learn
it all all over again, even if those notes may not be complete.


> Have a nice day :)

  :)  same to you!  :)


  songbird

[toc] | [prev] | [next] | [standalone]

#246006

Thomas Schmitt wrote:
> songbird wrote:
...
>> i have two versions of it now, but it was a five years ago and
>> i don't recall if i made the changes or got a new version from
>> the author.  :)
>
> Your request for a dry-run mode was fulfilled by a new version:
>   https://lists.debian.org/debian-user/2017/03/msg01268.html
> which needed a fix because of a misleading program message:
>   https://lists.debian.org/debian-user/2017/03/msg01270.html
> That's the version which is still available as
>   http://scdbackup.webframe.org/make_isombr_part.c
>   MD5 dd3e1a16e9593f908a1ce9ec848fd929
>
> (My local version is slightly younger and seems to have been augmented
> for being able to determine the USB stick size on FreeBSD, too. Dunno
> whether this was for real use or just some finger exercises for myself.)
>
>
>>   3. add boot loader (syslinux)
>
> Why this step ?
> The ISO brings a SYSLINUX/ISOLINUX boot loader for legacy BIOS (and GRUB
> for EFI). The MBR of the ISO hops onto the ISOLINUX boot image which then
> brings up the SYSLINUX boot menu.
>
> What negative effect did you see when not performing this step 3 ?

  now i've gone and checked to make sure i am using the same
version as above and i am so those edits were all done by you
and not me.  *whew!*

  i updated my notes to put a question next to that step so
that if i go to use the program again i'll get the answer to 
that question, but i have to say that i've now been through two
different machines since then and it may have been something
i did to get around some strange bios issue from back then
which may not apply any more.

  next time i get to a widget store i'll pick up an extra
USB stick and see what happens - they usually have them for
not much and i should always have an extra one around 
anyways.


  songbird

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.user

csiph-web