[Klug-general] RAID OF DEATH!!!

[Karl Lattimer]

Stuart Buckland wrote:

>On Thu, 2004-12-02 at 13:13 +0000, Karl Lattimer wrote:
>  
>
>>Please, anyone with some raid knowledge must be able to help!!! 
>>
>>I was testing my raid array, raid5 8x250Gb maxtor disks. all SATA, i
>>used raidsetfaulty /dev/mn0 /dev/sdd now that disk doesn't work at
>>all!!!!
>>
>>Does raidsetfaulty do something to the disk? Any help much appreciated.
>>
>>    
>>
>
>When you say 'doesn't work at all' do you mean it doesn't work in the
>array again or do you mean it's totally stopped functioning even as a
>single drive?
>
>If it's just stopped working in the array then you should be able to
>make it active again with raidhotadd <array> <drive> or something like
>that.
>
>Stu
>
>Stu
>
>
>
>_______________________________________________
>Kent mailing list
>Kent at mailman.lug.org.uk
>http://mailman.lug.org.uk/mailman/listinfo/kent
>
>  
>
It seems the fault lies with hot unplugging the wrong disk

The configuration is as follows

sata_via onboard the first chipset (only allows booting from here)
    channel 1   250Gb Maxtor Diamond Max 9 SATA 150   (hde)
       /boot    200Mb   ext2
       swap    700Mb
       Software RAID
    channel 2   250Gb Maxtor Diamond Max 9 SATA 150   (hdg)
       swap    700Mb
       Software RAID
sata_promise onboard the second chipset
    channel 1   250Gb Maxtor Diamond Max 9 SATA 150   (sda)
       Software RAID
    channel 2   250Gb Maxtor Diamond Max 9 SATA 150   (sdb)
       Software RAID
sata_promise maxtor pci third chipset
    channel 1   250Gb Maxtor Diamond Max 9 SATA 150   (sdc)
       Software RAID
    channel 2   250Gb Maxtor Diamond Max 9 SATA 150   (sdd)
       Software RAID
sata_promise maxtor pci forth chipset
    channel 1   250Gb Maxtor Diamond Max 9 SATA 150   (sde)
       Software RAID
    channel 2   250Gb Maxtor Diamond Max 9 SATA 150   (sdf)
       Software RAID

/dev/md0   ext3 1.6Tb

The arrangement in the case is thus

sdc      hde
sdd      hdg
sde      sda
sdf      sdb

no spares, all members. I'm new to this many disks on SATA so if 
anything seems dodgy there please let me know.

So here's what i think i've done...

I setfaulty sdd, hot removed sdd

then i counted down a,b,c,d along the front (stupidly) and pulled out 
the forth one. Yes I am embarrassed! Then in the panic that then 
followed as the system became unstable i unplugged the one above it 
because i thought that was sdd. A second mistake and we're still not 
finished yet.

The system then stopped, I rebooted. Nothing doing, I mean it wouldn't 
boot, even if i put the disks back in. Couldn't pivot to the raid device 
and therefore couldn't boot.

So I'm reinstalling. Now I'm getting major errors on formatting the raid 
again.

Thanks to gentoo (<-honorable mention) I have a system rescue CD which 
will hopefully give me the ability to return the disks to a working 
state, and _HOPEFULLY_ will give me the ability to format the raid once 
again.

!! WARNING !! Do not unplug disks without removing them from the array 
first, and having your head switched on and in gear and knowing which 
physical disk it is you want to unplug from the array.

I think i probably deserve 10 lashings with a cane for that

Thanks
    Karl,