PCLinuxOS-Forums
News: ...FLASH!!! ...New PCLinuxOS Testing board now open. Register today! Be an active contributor to the PCLinuxOS future! ... Read all about it now, on THIS forum!!!..
 
*
Welcome, Guest. Please login or register. May 24, 2012, 07:48:41 AM


Login with username, password and session length


Pages: [1] 2   Go Down
  Print  
Author Topic: Booting Flash Sticks & KDE Partition Manager  (Read 584 times)
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« on: January 01, 2012, 03:16:41 PM »

I only recently discovered this app in the repository.
It is 'almost' identical to Gparted in GUI etc etc.

I used it yesterday for the first time and was rather pleased with the results.

I had two out of three USB flash disks which would not boot on one PC.
I never could figure out why ......  rebuilt the Partition table etc with fdisk, Gparted and drakdisk ......
and thought there must be something wrong with the PC ......  although it would boot one of the devices which was the same model as one which it wouldn't.
All the devices would boot on other PCs I tried.

Yesterday I rebuilt the two 'problem' drives' Partition Tables using KDE Partition Manager, and created them as LiveUSb devices.
They now boot in the 'problem' PC!

So the question arises ......  why this should be so?
Apparently KDE Partition Manager did something differently to the other three apps, which the PC accepted.

In trying to figure out what could cause this, I recollect that when creating partitions this year, I set the alignment of the partitions to MiB rather than the older Cylinder alignment. This helps with the speed of access.

I *think* though am not yet sure, that KDE Partition Manager may use the older Cylinder alignment by default and this is why the BIOS of the PC will now accept those devices. (EDIT: See below ...  the version in the repo does default to Cylinder alignment it seems)

It is the only thing I can think of - at the moment - that might go some way to explaining what happened.

In any case I am grateful to have tested the KDE Partition Manager, as I now have two devices which will again boot on any PC I have tried - small sample as yet, but good so far.


If anyone would care to make comments on my thoughts above ......  how possible/impossible/likely/unlikely etc ......  I would be happy to read them.

It is rather intriguing ....

http://blog.volker-lanz.de/2010/05/30/new-in-kde-partition-manager-1-1-iii-support-for-4096-byte-sectors
Quote
Maximum Flexibility With KDE Partition Manager 1.1

KDE Partition Manager 1.0, like many partitioning tools in use today, still uses the traditional scheme for best compatibility with older MS-DOS based disk manipulation software. It will insist on aligning any partition it moves or resizes to a cylinder boundary. This has lead to some confusion among users already when the application moves a partition by a few sectors when making it larger or smaller.


As we still have version 1.0 in the repo, and I have no recollection of being asked for a decision about alignment, it seems as if the default to Cylinder alignment in the older KDE Partition Manager actually saved my day.

Note:  Yes I know there are no actual platters/cylinders/whatever in a flash drive  Cheesy

It would be interesting to hear is others have met with anything similar.

regards.

Extra:
             This is how the three look now .....  the first always booted, and the second and third are the two which were problematic. It is noticeable that the first is configured differently ......  Sad

Code:
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 16.0 GB, 16039018496 bytes
255 heads, 63 sectors/track, 1949 cylinders, total 31326208 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000760fd

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1            2048    26879999    13438976   83  Linux
/dev/sdm2        26880000    31068159     2094080   82  Linux swap / Solaris
/dev/sdm3        31068160    31326207      129024   83  Linux
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 16.0 GB, 16039018496 bytes
255 heads, 63 sectors/track, 1949 cylinders, total 31326208 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000419d4

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1              63      208844      104391   83  Linux
/dev/sdm2   *      208845    31310684    15550920   83  Linux
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 4139 MB, 4139778048 bytes
128 heads, 62 sectors/track, 1018 cylinders, total 8085504 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000c1175

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1              62      206335      103137   83  Linux
/dev/sdm2   *      206336     8078847     3936256   83  Linux
[user@XPS ~]$
Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
AS
Global Moderator
Hero Member
*****
Offline Offline

Posts: 4086

Have a nice ... night!


« Reply #1 on: January 01, 2012, 06:35:19 PM »

I only recently discovered this app in the repository.
It is 'almost' identical to Gparted in GUI etc etc.

I used it yesterday for the first time and was rather pleased with the results.

I had two out of three USB flash disks which would not boot on one PC.


Like our friend old-polack sometimes says, "you are telling us a story"  ... Cheesy Grin
What do you mean by "would not boot" ? BIOS error, grub error ? Black screen of death ? what ?

Quote
I never could figure out why ......  rebuilt the Partition table etc with fdisk, Gparted and drakdisk ......
and thought there must be something wrong with the PC ......  although it would boot one of the devices which was the same model as one which it wouldn't.


May be it's not important, but I agree here ... "there must be something wrong with the PC"

Quote
All the devices would boot on other PCs I tried.


To me it sound like the partition tables on each device are/were correct.

Quote
Yesterday I rebuilt the two 'problem' drives' Partition Tables using KDE Partition Manager, and created them as LiveUSb devices.
They now boot in the 'problem' PC!

So the question arises ......  why this should be so?
Apparently KDE Partition Manager did something differently to the other three apps, which the PC accepted.


I agree about the word "apparently"  Cheesy

Quote
In trying to figure out what could cause this, I recollect that when creating partitions this year, I set the alignment of the partitions to MiB rather than the older Cylinder alignment. This helps with the speed of access.

I *think* though am not yet sure, that KDE Partition Manager may use the older Cylinder alignment by default and this is why the BIOS of the PC will now accept those devices. (EDIT: See below ...  the version in the repo does default to Cylinder alignment it seems)


Alignment/Misalignment can affect performance for sure, ... but cannot affect partition boundaries recognition, instead, a block size different from the historical 512 byte per sector (like in the case of 4KB sector size) can. Partition tables are made using sector numbers, they will be different depending on sector size. As far as I know actually the sector size (physical, logical, optimal) comes always from the device itself, not from the partitioning tool.

All your devices shown a sector size of 512 bytes.

Quote
It is the only thing I can think of - at the moment - that might go some way to explaining what happened.


Where do you installed Grub ? MBR ? EBR ?  Lot of chance of alternative issues around here ...  Grin

Quote
In any case I am grateful to have tested the KDE Partition Manager, as I now have two devices which will again boot on any PC I have tried - small sample as yet, but good so far.


Out of curiosity, had installed kde partition manager, at first use I was unable to format an empty USB stick (completely empty, filled out with random data from /dev/urandom ...), but was able to manage it after adding an empty partition table using fdisk.

Results may vary  Cheesy

Quote
If anyone would care to make comments on my thoughts above ......  how possible/impossible/likely/unlikely etc ......  I would be happy to read them.

It is rather intriguing ....

Indeed!  Grin

Quote
http://blog.volker-lanz.de/2010/05/30/new-in-kde-partition-manager-1-1-iii-support-for-4096-byte-sectors
Quote
Maximum Flexibility With KDE Partition Manager 1.1

KDE Partition Manager 1.0, like many partitioning tools in use today, still uses the traditional scheme for best compatibility with older MS-DOS based disk manipulation software. It will insist on aligning any partition it moves or resizes to a cylinder boundary. This has lead to some confusion among users already when the application moves a partition by a few sectors when making it larger or smaller.


As we still have version 1.0 in the repo, and I have no recollection of being asked for a decision about alignment, it seems as if the default to Cylinder alignment in the older KDE Partition Manager actually saved my day.

Note:  Yes I know there are no actual platters/cylinders/whatever in a flash drive  Cheesy

It would be interesting to hear is others have met with anything similar.

regards.

Extra:
             This is how the three look now .....  the first always booted, and the second and third are the two which were problematic. It is noticeable that the first is configured differently ......  Sad

Code:
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 16.0 GB, 16039018496 bytes
255 heads, 63 sectors/track, 1949 cylinders, total 31326208 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000760fd

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1            2048    26879999    13438976   83  Linux
/dev/sdm2        26880000    31068159     2094080   82  Linux swap / Solaris
/dev/sdm3        31068160    31326207      129024   83  Linux
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 16.0 GB, 16039018496 bytes
255 heads, 63 sectors/track, 1949 cylinders, total 31326208 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000419d4

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1              63      208844      104391   83  Linux
/dev/sdm2   *      208845    31310684    15550920   83  Linux
[user@XPS ~]$ fdisk -l /dev/sdm

Disk /dev/sdm: 4139 MB, 4139778048 bytes
128 heads, 62 sectors/track, 1018 cylinders, total 8085504 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000c1175

   Device Boot      Start         End      Blocks   Id  System
/dev/sdm1              62      206335      103137   83  Linux
/dev/sdm2   *      206336     8078847     3936256   83  Linux
[user@XPS ~]$


One thing I noticed is the boot flag, absent on the first device, active on the second partition on each of the other two devices.  Roll Eyes

AS
Logged
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« Reply #2 on: January 01, 2012, 07:25:57 PM »

Re the story ......  Cheesy .....  of the two that would not boot ......  one (the 4GB) prevented the BIOS from going through POST. Just stopped dead and needed a hard power off to get back control.

The other ...  the second in the list  ....  did not prevent the BIOS from operating, but was never presented as a bootable option. It was simply ignored.

The first was always recognised and booted if selected.

Boot Flag:  The first has none set and has always booted.

"there must be something wrong with the PC":  Rather than something wrong, I am now inclined to the view that there is some limitation in the BIOS. (Dell version 1.08 I think, but can check if required)

Where do you installed Grub ?:  All devices have Grub installed to the MBR, as it is a necessary requirement to boot the device. As I said all the devices booted in other machines.


Quote
Out of curiosity, had installed kde partition manager, at first use I was unable to format an empty USB stick (completely empty, filled out with random data from /dev/urandom ...), but was able to manage it after adding an empty partition table using fdisk.

That too is a story  Cheesy ......  format an empty deviceCheesy   Did you try creating a new partition table with KDE Partition Manager?

So, no nearer an explanation of what might possibly have caused it.  Sad

All I can think of is a BIOS limitation regarding the recognition of drives that have other than Cylinder alignment on their partitions.

regards.
Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
AS
Global Moderator
Hero Member
*****
Offline Offline

Posts: 4086

Have a nice ... night!


« Reply #3 on: January 01, 2012, 07:48:18 PM »

Quote
Out of curiosity, had installed kde partition manager, at first use I was unable to format an empty USB stick (completely empty, filled out with random data from /dev/urandom ...), but was able to manage it after adding an empty partition table using fdisk.

That too is a story  Cheesy ......  format an empty device?  Cheesy   Did you try creating a new partition table with KDE Partition Manager?

Sorry, I have used the wrong words here, meant to say "create a new partition table" ... all buttons were greyed out, no action was possible.

Quote
So, no nearer an explanation of what might possibly have caused it.  Sad

Maya profecy ?  Cheesy Grin

Quote
All I can think of is a BIOS limitation regarding the recognition of drives that have other than Cylinder alignment on their partitions.

regards.

Can you replicate the issue ? if yes, a dump of the first 512 bytes would be interesting to compare against the the same of a working device ...


EDIT: However it is possible to format a not partitioned device, simply use mkfs.ext2 /dev/sdc, it will not be a bootable device but will be perfectly usable as a storage device.
Logged
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« Reply #4 on: January 01, 2012, 08:02:17 PM »

Quote
Can you replicate the issue ? if yes, a dump of the first 512 bytes would be interesting to compare against the the same of a working device ...

I am hesitant to try ...... if the Cylinder alignment is not the issue then I could be left back with a dud again  Sad

Maybe when I am finished with the current version of LiveUSB I might try it ....  not 'till then though ....
Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
AS
Global Moderator
Hero Member
*****
Offline Offline

Posts: 4086

Have a nice ... night!


« Reply #5 on: January 01, 2012, 08:19:04 PM »

if the Cylinder alignment is not the issue then I could be left back with a dud again  Sad


may be ... apparently some BIOS may be affected from Cylinders misalignment,
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=614632    message #14
Quote
The only reason you might still need cylinder alignment is if you have a
buggy BIOS that gets confused by non-cylinder-aligned partitions


 Wink
Logged
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« Reply #6 on: January 01, 2012, 08:35:21 PM »

if the Cylinder alignment is not the issue then I could be left back with a dud again  Sad


may be ... apparently some BIOS may be affected from Cylinders misalignment,
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=614632    message #14
Quote
The only reason you might still need cylinder alignment is if you have a
buggy BIOS that gets confused by non-cylinder-aligned partitions


 Wink


AH!!!!!!!!!

Thank you!

As it is a Dell BIOS I am prepared to believe anything about it!  Grin

I will feel more inclined to try a new partition table with MiB alignment after reading that .......  when I can afford to be without the USB stick for a little while.

Thanks for finding that, AS ..... it is somewhat comforting to read that my thoughts were not as weird as they appeared at first  Cheesy  Cheesy  Grin
Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
AS
Global Moderator
Hero Member
*****
Offline Offline

Posts: 4086

Have a nice ... night!


« Reply #7 on: January 01, 2012, 09:08:06 PM »


Thanks for finding that, AS ..... it is somewhat comforting to read that my thoughts were not as weird as they appeared at first  Cheesy  Cheesy  Grin

I know that feeling ... when you are reporting something .. and people look at you like at an alien  Wink Cheesy

meanwhile, at opposite side ...

A woman, friend of my wife, once call her on the phone: "My computers is died .. it doesn't turn on anymore ..."
my wife: "Did you checked the plug cord ? "
friend: "Ohhh .. you are an Hacker!!!"     Cheesy Cheesy Grin Grin

 Wink
Logged
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« Reply #8 on: January 01, 2012, 09:14:32 PM »

Cheesy  Cheesy  Cheesy
Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
Vortеx
Hero Member
*****
Offline Offline

Posts: 2010


Vorteks without the Secretary


« Reply #9 on: January 02, 2012, 04:23:37 AM »

Somehow the LiveUSB partition of the flash stick got written to my /etc/fstab on the hard disk.
Is it supposed to be this way?
Logged
AS
Global Moderator
Hero Member
*****
Offline Offline

Posts: 4086

Have a nice ... night!


« Reply #10 on: January 02, 2012, 05:45:02 AM »

Somehow the LiveUSB partition of the flash stick got written to my /etc/fstab on the hard disk.
Is it supposed to be this way?

No, probably is the result of formatting the USB stick using Control Center and setting up a mount point there.  Wink

AS
Logged
Vortеx
Hero Member
*****
Offline Offline

Posts: 2010


Vorteks without the Secretary


« Reply #11 on: January 02, 2012, 06:18:17 AM »

No.
It was a remnant in fstab of the hard disk that I recently detached.
The labels were equal and the system threw warnings the flash partition is listed in the fstab.
Noob here you know Wink
Logged
melodie
Hero Member
*****
Offline Offline

Posts: 6007


XMMP=Jabber, free instant messaging protocol !


WWW
« Reply #12 on: January 02, 2012, 09:02:37 PM »

Hi,
I don't know the K partition manager, but it must be a good one, I read something about it at Wikipedia once, I think. (http://en.wikipedia.org/wiki/KDE_Partition_Manager)

I almost always use Gparted, and never paid attention to the default choice : I just looked, as I wanted to do some changes on a hard drive it is MiB, not cylinder when a partition is created.

I wonder if this is not the point about a problem I met twice in a few years time : formatting a partition with drakdisk (once in pclos a pair of years ago, and once in Mandriva I think, more years ago yet). I had all partitions confused and could not boot to any after (I was multibooting, and got message about a partition which was overlapping). I remember the second time, searching during several days for a formatting tool that would be able to see my partitions, then I erased the swap to free an overlapped cylinder, (found one in a very old ash shell on some live) and the problem was solved.

Perhaps one was using the method with cylinders and the other MiB, then the last partition table was not matching perfectly the one under ?

Logged

melodie at swissjabber dot ch - IRC #pclinuxos-fr sur freenode
wedgetail
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 2110


Any Bugs in site?


« Reply #13 on: February 04, 2012, 06:38:30 AM »

Just18

I will assume that you know but for somebody landing in this topic.  I will just add that the fdisk results on your 3 USB keys you mention the first one, 16 GB unit which shows a starting sector of 2048. This is likely because the USB Key has been formatted with fdisk new (To cope with bigger drives)

In case one has an older say 2009 liveCD then formatting would be using fdisk-old starting sector number likely to be under 100.  I am not clear on why your two others 'came good' but doing some calculations. (Of course against Maurice's  Grin)

Is there any chance you have a record of the non working partition tables?

Logged

KDE & KDE-mini, ASUSTek P5P41D Rev X.0x, BIOS AMI0207 07/21/2009, "Pentium(R) Dual-Core CPU E5300 @ 2.60GHz", nVidia GeForce 9600 GT, 2x1GB Seagate Technology 1000528AS HDD
TV CompuPro VideoMate Vista E700 (not working in Linux), Acer X243HD LCD Screen
Just18
PCLinuxOS Tester
Hero Member
*******
Offline Offline

Posts: 4591


MLUs Forever!


« Reply #14 on: February 04, 2012, 07:02:54 AM »

Just18

I will assume that you know but for somebody landing in this topic.  I will just add that the fdisk results on your 3 USB keys you mention the first one, 16 GB unit which shows a starting sector of 2048. This is likely because the USB Key has been formatted with fdisk new (To cope with bigger drives)

In case one has an older say 2009 liveCD then formatting would be using fdisk-old starting sector number likely to be under 100.  I am not clear on why your two others 'came good' but doing some calculations. (Of course against Maurice's  Grin)

Is there any chance you have a record of the non working partition tables?



Not a chance unfortunately.
It was a problem that existed for some time ......  in hindsight, probably from the recreation of the partitions in the new manner.
I did not get the urge to go back and redo the structure of the USB stick .....  say using fdisk or Gparted  ....  to confirm that this was the problem. I might do so in the future ....  but there would be little benefit to me and it just *might* not come good again if I am wrong  Cheesy

Logged

MLUs rule the roost!

Linux XPS 3.2.17-pclos1.pae.bfs  32 bit
Intel(R) Core(TM)2 Quad  CPU   Q9450  @ 2.66GHz
4 GB RAM
MCP51 High Def Audio
GeForce GTX 550 Ti
PHILIPS  ‎DVD+-RW DVD8701
‎Logitech ‎BT Mini-Receiver
Afatech ‎DVB-T 2 USB DTT
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.16 | SMF © 2011, Simple Machines

Valid XHTML 1.0! Valid CSS! Dilber MC Theme by HarzeM