First, let me say thank you very, very much for your help, Håkon. I appreciate the time you took to reply very, very much.
> If you had just one partition that filled the whole disk, this should be > not difficult to recreate. I agree, however I would like to point out that the partition hasn't actually been destroyed -- just horribly mangled. However I acknowledge the very high likelyhood that this partition will have to be destroyed in order to create a new, proper one -- hopefully then it can be recovered accordingly. > To recreate a partition that have gone missing or been destroyed it has to > start at the exact same sector that it was originally created. The original partition simply consumed the whole disc from end to end, so I shouldn't have to rack my brain to figure this out. > The start/end numbers above are not exact sector numbers so I would like to > have some more information. Could you run parted and print the complete > partition table using units s, cyl and chs? Absolutely: ------- sectors ------- Model: AMCC 9550SX-12M DISK (scsi) Disk /dev/sdb: 7812415488s Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start End Size Type File system Flags 1 63s 3517442203s 3517442141s primary xfs --------- cylinders --------- Disk /dev/sdb: 486300cyl Sector size (logical/physical): 512B/512B BIOS cylinder,head,sector geometry: 486300,255,63. Each cylinder is 8225kB. Partition Table: msdos Number Start End Size Type File system Flags 1 0cyl 218950cyl 218950cyl primary xfs ----------------------- cylinders/heads/sectors ----------------------- Disk /dev/sdb: 486300,95,2 Sector size (logical/physical): 512B/512B BIOS cylinder,head,sector geometry: 486300,255,63. Each cylinder is 8225kB. Partition Table: msdos Number Start End Type File system Flags 1 0,1,0 218950,165,58 primary xfs > Could you run fdisk with commands "x" and "p" to print the raw partition? Yes, here it is: Disk /dev/sdb: 255 heads, 63 sectors, 486300 cylinders Nr AF Hd Sec Cyl Hd Sec Cyl Start Size ID 1 00 1 1 0 254 63 1023 63 3517442141 83 2 00 0 0 0 0 0 0 0 0 00 3 00 0 0 0 0 0 0 0 0 00 4 00 0 0 0 0 0 0 0 0 00 I have to assert that I really don't know what I'm doing with advanced partitioning stuff; fdisk is all I have really ever used until I discovered it wouldn't handle partitions in excess of 2TB. In light of this fact, I have to disclose a couple extra points that may make all the difference in the world: 1) I am fairly certain, but NOT dead positive, that I had set the disk label to "gpt" when I created the array, following a brief guide found elsewhere on the subject. This therefore begs the question of why the disk label is showing as 'msdos' in I am terrified that this post may apply to me: http://www.mail-archive.com/bug-parted@gnu.org/msg02465.html 2) I am ashamed to say CONFIG_EFI_PARTITION=y was NOT set in the kernel when I reboot the machine. It is defined now and the box has been rebooted (but it makes no difference to the detected partition size etc.). In light of these facts I'm wondering if it isn't possible that I created a 4TB partition in parted using an msdos table (despite the fact that this is effectively illegal, which is the point of the link above). I'm not precisely sure how this equates to data corruption, however, because as I already stated the device ran fine for about a month (with no reboots) and there were no cases of data corruption on the ~250GB or so data that was on the drive (although I don't doubt if I had gone well over that amount there would be corruption problems). This problem only became apparent after a reboot, and it goes without saying that a couple successive reboots have done little to improve the situation. :) Thank you again for taking time out of your Sunday to help; I am grateful. Cheers, Kevin Williams _______________________________________________ bug-parted mailing list bug-parted@gnu.org http://lists.gnu.org/mailman/listinfo/bug-parted
