slackware -current softraid5 boot problem

slackware -current softraid5 boot problem

[Dexter Filmore]

Finally got me a bunch of disks as a raid 5, I don't even want to boot from 
it, plain data array. four sata-II samsungs on a sil3114 controller, all fine 
so far.

Booting the machine I get "no superblock on /dev/sdd, stopping 
assembly" (or assembly aborted or so)
After boot has completed, running mdadm --assemble --scan --verbose manually 
works just fine and doesn't tell a thing about missing superblocks. 

Ideas, anyone?

Dex


-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS d--(+)@ s-:+ a- C+++(++++) UL+>++++ P+>++ L+++>++++ E-- W++ N o? K-
w--(---) !O M+ V- PS++(+) PE(-) Y++ PGP t++(---)@ 5 X+(++) R+(++) tv--(+)@ 
b++(+++) DI+++ D G++ e* h>++ r%>* y?
------END GEEK CODE BLOCK------

http://www.stop1984.com
http://www.againsttcpa.com

Re: slackware -current softraid5 boot problem

[Luca Berra]

On Tue, May 09, 2006 at 12:06:32AM +0200, Dexter Filmore wrote:
>Finally got me a bunch of disks as a raid 5, I don't even want to boot from 
>it, plain data array. four sata-II samsungs on a sil3114 controller, all fine 
>so far.
>
>Booting the machine I get "no superblock on /dev/sdd, stopping 
>assembly" (or assembly aborted or so)
>After boot has completed, running mdadm --assemble --scan --verbose manually 
>works just fine and doesn't tell a thing about missing superblocks. 
>
you don't give a lot of information about your setup,
in any case it could be something like udev and the /dev/sdd device node
not being available at boot?

Regards,
L.

-- 
Luca Berra -- bluca@comedia.it
        Communication Media & Services S.r.l.
 /"\
 \ /     ASCII RIBBON CAMPAIGN
  X        AGAINST HTML MAIL
 / \

Re: slackware -current softraid5 boot problem - additional info

[Dexter Filmore]

Am Dienstag, 9. Mai 2006 07:50 schrieb Luca Berra:
> you don't give a lot of information about your setup,

You're sure right here, I was a bit off track yesterday from tinkering till 
night - info below.

> in any case it could be something like udev and the /dev/sdd device node
> not being available at boot?

Ok: 
Slackware-current with kernel 2.6.14.6, *no* udev, plain old hotplug
I had to put the raid start script in a reasonable place myself (not preconfed 
in Slack) so I have to figure yet if sees /etc/mdadm.conf  when the script is 
called. (If presence of mdadm.conf is totally uninteresting, let me know, I 
just started on raid.)
The other disks are seen fine, and since they are all the same type on the 
same controller there's no reason why it is not seen then.
(Unless for some reason mdadm talks to the *last* disk first and then stops - 
else it should complain about sda rather.)


***** mdadm -E info *****


# mdadm -E /dev/sdd
/dev/sdd:
          Magic : a92b4efc
        Version : 00.90.02
           UUID : db7e5b65:e35c69dc:7c267a5a:e676c929
  Creation Time : Mon May  8 00:05:16 2006
     Raid Level : raid5
    Device Size : 244198464 (232.89 GiB 250.06 GB)
     Array Size : 732595392 (698.66 GiB 750.18 GB)
   Raid Devices : 4
  Total Devices : 4
Preferred Minor : 0

    Update Time : Tue May  9 00:43:46 2006
          State : clean
 Active Devices : 4
Working Devices : 4
 Failed Devices : 0
  Spare Devices : 0
       Checksum : 61f0ffd6 - correct
         Events : 0.24796

         Layout : left-symmetric
     Chunk Size : 32K

      Number   Major   Minor   RaidDevice State
this     3       8       48        3      active sync   /dev/sdd

   0     0       8        0        0      active sync   /dev/sda
   1     1       8       16        1      active sync   /dev/sdb
   2     2       8       32        2      active sync   /dev/sdc
   3     3       8       48        3      active sync   /dev/sdd

***** mdstat *****

Once I started the array manually (which works fine) mdstats look like:

# cat /proc/mdstat
Personalities : [raid5]
md0 : active raid5 sda1[0] sdd1[3] sdc1[2] sdb1[1]
      732563712 blocks level 5, 32k chunk, algorithm 2 [4/4] [UUUU]

unused devices: <none>

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS d--(+)@ s-:+ a- C+++(++++) UL+>++++ P+>++ L+++>++++ E-- W++ N o? K-
w--(---) !O M+ V- PS++(+) PE(-) Y++ PGP t++(---)@ 5 X+(++) R+(++) tv--(+)@ 
b++(+++) DI+++ D G++ e* h>++ r%>* y?
------END GEEK CODE BLOCK------

http://www.stop1984.com
http://www.againsttcpa.com

Re: slackware -current softraid5 boot problem - additional info

[Mike Hardy]

Something fishy here

Dexter Filmore wrote:

> # mdadm -E /dev/sdd

Device /dev/sdd

> # cat /proc/mdstat
> Personalities : [raid5]
> md0 : active raid5 sda1[0] sdd1[3] sdc1[2] sdb1[1]
>       732563712 blocks level 5, 32k chunk, algorithm 2 [4/4] [UUUU]

Components that are all the first partition.

Are you using the whole disk, or the first partition?

It appears that to some extent, you are using both.

Perhaps some confusion on that point between your boot scripts and your
manual run explains things?


-Mike

Re: softraid5 boot problem - partly my fault, solved

[Dexter Filmore]

Mystery solved: had to probe another module. 
Wait, wait, I can defend myself :)

What led me to believe the controller was autoprobed during boot is that mdadm 
complained about *sdd*, but not about sd[abc], hence I assumed [abc] were all 
fine.
Plus, I didn't have to probe the module manually after boot was completed 
(appears that at that point some other module inserted it as a dependency).

So - is that how mdadm (or the kernel?) handle raid, is the last disk checked 
first by design?

Dex

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS d--(+)@ s-:+ a- C+++(++++) UL+>++++ P+>++ L+++>++++ E-- W++ N o? K-
w--(---) !O M+ V- PS++(+) PE(-) Y++ PGP t++(---)@ 5 X+(++) R+(++) tv--(+)@ 
b++(+++) DI+++ D G++ e* h>++ r%>* y?
------END GEEK CODE BLOCK------

http://www.stop1984.com
http://www.againsttcpa.com