From: stryder@telenet.be.invalid   
      
   On Tuesday 18 September 2012 14:18, Fred conveyed the following to   
   alt.os.linux.mandriva...   
      
   > Hi all,   
   > this is no mandriva system but busybox NAS.   
   > But as it's a RAID problem, i ask here knowing that plenty of people   
   > will be kind to help me ;-)   
   >   
   > The problem is that volume groups have disapeared suddenly :   
   > vgdisplay ==>  No volume groups found   
   > lvdisplay ==>  No volume groups found   
   > pvdisplay returns nothing   
      
   Have you run vgscan, possibly with the "--mknodes" option?   
      
   In a normal GNU/Linux distribution, vgscan is normally executed from   
   within the init scripts by default, but I don't know whether busybox-   
   based systems do that.   
      
   > df -h gives :   
   >   
   > /dev/md0                  2.1M      2.1M         0 100% /old   
   > /dev/md0                197.4M     77.3M    110.0M  41% /   
   > /dev/md2                  2.7T    120.6G      2.6T   4% /DataVolume   
   > /dev/ram0                61.6M      5.5M     56.2M   9% /mnt/ram   
   > /dev/md2                  2.7T    120.6G      2.6T   4%   
   > /shares/Download   
   > /dev/md2                  2.7T    120.6G      2.6T   4% /shares/Public   
   > ................list continued with all the other shares   
   >   
   > ######### So DATAVOLUME is there !!! #############   
   >   
   >   mdadm -E /dev/sda4 returns :   
   > /dev/sda4:   
   >            Magic : a92b4efc   
   >          Version : 0.90.00   
   >             UUID : c9d6f3c0:76a8d365:44bb6e8c:b20b82e6   
   >    Creation Time : Wed Jun 15 15:06:59 2011   
   >       Raid Level : raid5   
   >    Used Dev Size : 975146112 (929.97 GiB 998.55 GB)   
   >       Array Size : 2925438336 (2789.92 GiB 2995.65 GB)   
   >     Raid Devices : 4   
   >    Total Devices : 4   
   > Preferred Minor : 2   
   >   
   >      Update Time : Mon Sep 10 11:20:44 2012   
   >            State : clean   
   >   Active Devices : 4   
   > Working Devices : 4   
   >   Failed Devices : 0   
   >    Spare Devices : 0   
   >         Checksum : b8ee4d79 - correct   
   >           Events : 699490   
   >   
   >           Layout : left-symmetric   
   >       Chunk Size : 64K   
   >   
   >        Number   Major   Minor   RaidDevice State   
   > this     3       8        4        3      active sync   /dev/sda4   
   >   
   >     0     0       8       52        0      active sync   /dev/sdd4   
   >     1     1       8       36        1      active sync   /dev/sdc4   
   >     2     2       8       20        2      active sync   /dev/sdb4   
   >     3     3       8        4        3      active sync   /dev/sda4   
   >   
   > ######### So RAID5 seems clean #############"   
   >   
   > BTW : i don't have the "datavolume doesn't exist" message, and   
   > absolutely no error show in the logs.   
   >   
   > do you think my RAID is still active ?   
      
   I'm not an expert on software RAID, but the "active sync" does seem to   
   suggest so.   
      
   > The NAS is in production in a professionnal environment and i would   
   > like to solve the problem without having to reset to factory or   
   > reconstruct the raid, the shares.....   
   >   
   > Knowing that all the shares are still accessible via network, may i   
   > safely try to reconstruct the VGs, LVs... ? Won't this rebuild the   
   > raid (20 hours), forcing me to restore all my data after recreating   
   > all the shares ?   
      
   Could be that all you need is to scan the drives for volume groups.   
      
   --   
   = Aragorn =   
   (registered GNU/Linux user #223157)   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)