0

for some reason one of my disks doesn't show its partitions in /dev/, however when I check it with cfdisk they are there, and writing the partition table to it again makes them reappear.

This is kind of annoying as I want to use it in a raid config, but the raid config can't find it so I have to manually add them on boot causing them to be resynced all over again.

OS: Debian 5.0 amd64 with backport kernel for ext4 support Disks: 2x Samsung Spinpoint F3 HD103SJ, 1x old IDE disk Other info is probably more or less irrelevant, if you need any command output I can supply it.

vmhost:/home/xen# fdisk -l

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000d6074

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1        6079    48829536   fd  Linux raid autodetect
/dev/sda2            6080      121601   927930465   fd  Linux raid autodetect

Disk /dev/sdb: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000cdeb8

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               1        6079    48829536   fd  Linux raid autodetect
/dev/sdb2            6080      121601   927930465   fd  Linux raid autodetect

Disk /dev/sdc: 250.0 GB, 250058268160 bytes
255 heads, 63 sectors/track, 30401 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xdba2dba2

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1   *           1       29164   234259798+  83  Linux
/dev/sdc2           29165       30401     9936202+   5  Extended
/dev/sdc5           29165       30401     9936171   82  Linux swap / Solaris

Disk /dev/md0: 50.0 GB, 50001346560 bytes
2 heads, 4 sectors/track, 12207360 cylinders
Units = cylinders of 8 * 512 = 4096 bytes
Disk identifier: 0x00000000

Disk /dev/md0 doesn't contain a valid partition table

Disk /dev/md1: 950.2 GB, 950200696832 bytes
2 heads, 4 sectors/track, 231982592 cylinders
Units = cylinders of 8 * 512 = 4096 bytes
Disk identifier: 0x00000000

Disk /dev/md1 doesn't contain a valid partition table
Not Available
  • 226
  • 1
  • 16

2 Answers2

1

If you are talking about 

Disk /dev/md0 doesn't contain a valid partition table

then it is expected output in any raid device.

In case you mean that devices with name '/dev/sda1', etc. do not exist in '/dev' then try using 

partprobe /dev/sda

when the devices are not listed. If this solves the problem, then you can add 'partprobe /dev/sda' to '/etc/rc.d/rc.local' file. This would still cause problem if you try to mount '/dev/sda1' etc. on boot using '/etc/fstab'

Saurabh Barjatiya
  • 4,703
  • 2
  • 30
  • 34
  • Exactly, in most cases you *don't* want a partition table on a RAID device, Linux just allows for that and fdisk expects this on block devices it opens. If you want to further split your RAID use LVM. – Hubert Kario Nov 07 '10 at 09:18
  • The problem was with I think with sda, every time I booted only /dev/sda existed while fdisk (As can be seen above) did detect /dev/sda1 and /dev/sda2. When writing the partition table to the disk again after this they would appear, until I rebooted. – Not Available Nov 08 '10 at 08:10
0

Stupid me apparently created a software raid superblock on /dev/sda a while ago which caused the software raid tool to mangle the partition data or something.

Not Available
  • 226
  • 1
  • 16