qtparted and kernel disagree about partitions
in [Debian]
|
Prev: OSSIM
Next: bigmem kernel
From: David Goodenough on 4 Mar 2010 10:50 I have a hard drive with a primary partition and an extended partiton. QTParted says that it has /dev/hda1, /dev/hda2 and /dev/hda5 which is what one would expect, but when I boot the disk the kernel reports that there are only the /dev/hda1 and /dev/hda2 partitons, no /dev/hda5. The really odd thing is that qtparted is running on the kernel which can not see /dev/hda5. The disk is question is a 2.5" 160GB drive, and is quite new. The kernel log shows no errors. Any idea what might cause this and how to fix it (preferably without repartitioning the disk as backing it all up would take a while and also I do not know how I would access the data in /dev/hda5 if I can not see it). David -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/201003041544.10116.david.goodenough(a)btconnect.com
From: Jordan Metzmeier on 4 Mar 2010 11:00 On Thu, Mar 4, 2010 at 10:44 AM, David Goodenough <david.goodenough(a)btconnect.com> wrote: > I have a hard drive with a primary partition and an extended partiton. > QTParted says that it has /dev/hda1, /dev/hda2 and /dev/hda5 which is > what one would expect, but when I boot the disk the kernel reports > that there are only the /dev/hda1 and /dev/hda2 partitons, no /dev/hda5. > > The really odd thing is that qtparted is running on the kernel which > can not see /dev/hda5. > > The disk is question is a 2.5" 160GB drive, and is quite new. The kernel > log shows no errors. > > Any idea what might cause this and how to fix it (preferably without > repartitioning the disk as backing it all up would take a while and also > I do not know how I would access the data in /dev/hda5 if I can not see > it). > > David > What version of Debian is this? What kernel version? Storage controller type? I find it odd that your disks would be named hdx as modern libata should emulate those drives as scsi. -- Jordan Metzmeier -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/50e5edd51003040754t3c6bdc4ai10feb575e7f666f5(a)mail.gmail.com
From: David Goodenough on 4 Mar 2010 11:10 On Thursday 04 March 2010, Jordan Metzmeier wrote: > On Thu, Mar 4, 2010 at 10:44 AM, David Goodenough > > <david.goodenough(a)btconnect.com> wrote: > > I have a hard drive with a primary partition and an extended partiton. > > QTParted says that it has /dev/hda1, /dev/hda2 and /dev/hda5 which is > > what one would expect, but when I boot the disk the kernel reports > > that there are only the /dev/hda1 and /dev/hda2 partitons, no /dev/hda5. > > > > The really odd thing is that qtparted is running on the kernel which > > can not see /dev/hda5. > > > > The disk is question is a 2.5" 160GB drive, and is quite new. The kernel > > log shows no errors. > > > > Any idea what might cause this and how to fix it (preferably without > > repartitioning the disk as backing it all up would take a while and also > > I do not know how I would access the data in /dev/hda5 if I can not see > > it). > > > > David > > What version of Debian is this? What kernel version? Storage controller > type? > > I find it odd that your disks would be named hdx as modern libata > should emulate those drives as scsi. > This is sid, running 2.6.32-2, and the controller is an Intel 82801DBM (ICH4-M) IDE Controller (rev 03). I thought that the libata change was coming in in 2.6.33, but maybe I misread something. The file system was built a little while ago, back in 2.6.26 or earlier days but I can not go back to 2.6.26 as udev would not work. David -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/201003041603.15609.david.goodenough(a)btconnect.com
From: Stephen Powell on 4 Mar 2010 11:50 On Thu, 4 Mar 2010 10:44:10 -0500 (EST), David Goodenough wrote: > > I have a hard drive with a primary partition and an extended partiton. > QTParted says that it has /dev/hda1, /dev/hda2 and /dev/hda5 which is > what one would expect, but when I boot the disk the kernel reports > that there are only the /dev/hda1 and /dev/hda2 partitons, no /dev/hda5. What happened to /dev/hda3 and /dev/hda4? How can there be a /dev/hda5 without all the partitions in between? > The really odd thing is that qtparted is running on the kernel which > can not see /dev/hda5. Perhaps qtparted reads the partition table directly, rather than relying on the kernel for this information. > Any idea what might cause this and how to fix it (preferably without > repartitioning the disk as backing it all up would take a while and also > I do not know how I would access the data in /dev/hda5 if I can not see > it). This is pure guesswork, but maybe the kernel stopped when it couldn't find /dev/hda3 and assumed that there was nothing beyond. Maybe creating dummy partitions /dev/hda3 and /dev/hda4 would allow the kernel to find /dev/hda5. It's worth a shot. I'm not even addressing the issue of SCSI vs ATA. I'll leave that to those who know more about it. (David, please excuse the duplicate. I accidentally replied to you personally instead of to the list.) -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/134597558.16750431267720927300.JavaMail.root(a)md01.wow.synacor.com
From: Boyd Stephen Smith Jr. on 4 Mar 2010 12:00
In <134597558.16750431267720927300.JavaMail.root(a)md01.wow.synacor.com>, Stephen Powell wrote: >On Thu, 4 Mar 2010 10:44:10 -0500 (EST), David Goodenough wrote: >> I have a hard drive with a primary partition and an extended partiton. >> QTParted says that it has /dev/hda1, /dev/hda2 and /dev/hda5 which is >> what one would expect, but when I boot the disk the kernel reports >> that there are only the /dev/hda1 and /dev/hda2 partitons, no /dev/hda5. > >What happened to /dev/hda3 and /dev/hda4? How can there be a /dev/hda5 >without all the partitions in between? Using "MS-DOS" partitioning: 1-4 are the Primary partitions. Their extents are recorded on sector 0 of the drive. 5-15 are the Logical partitions. Their extents are recorded in a partition table store inside primary partition (which is not otherwise used). bss(a)monster:~% ls /dev/sd[fg]* /dev/sdf /dev/sdf5 /dev/sdg /dev/sdg5 /dev/sdf1 /dev/sdf6 /dev/sdg1 /dev/sdg6 >> The really odd thing is that qtparted is running on the kernel which >> can not see /dev/hda5. > >Perhaps qtparted reads the partition table directly, rather than >relying on the kernel for this information. Yes, it does. It's also possible that the kernel doesn't understand the partition table, but qtparted does. I've never had this happen, but it might be useful to see any kernel messages about hda, particularly about partitioning. These messages would appear in /var/log/kern.log, I think. The output of (parted -l) might also be useful. It's possible there's a inconsistency in your partition table that qtparted corrects but the kernel does not. >> I do not know how I would access the data in /dev/hda5 if I can not see >> it). device-mapper or losetup using /dev/hda as the basis device and a rather large, manually calculated offset. If you just want the bytes backed up, just dd from /dev/hda. -- Boyd Stephen Smith Jr. ,= ,-_-. =. bss(a)iguanasuicide.net ((_/)o o(\_)) ICQ: 514984 YM/AIM: DaTwinkDaddy `-'(. .)`-' http://iguanasuicide.net/ \_/ |