Re: Vinum dumpconfig returns nothing ?!?

From: Vincent Rioux <vrioux_at_centrenad.com>
Date: Fri, 13 Feb 2004 22:30:12 -0500 (EST)
Whoa Greg, now you really did put me back on track!  In one single mail
you answered all my questions.  A little googling for Disklabel and I was
all set.  Now all is working fine!!!

Just in case someone else runs into the same kind of problem as I did,
disklabel can be a pain in the ass if you don't know very well what it
does.  So if you mess up your partition tables, and disklabel refuses to
write a new config to the disk, use :

dd if=/dev/zero of=/dev/rad1 bs=1k count=1

(replacing rad1 by the raw device name of your disk).  This zeroes the
first two sectors of the drive, enabling disklabel to write the config
there once again.

Thanks again for the outstanding help.  :)

Vincent

>
> On Friday, 13 February 2004 at 10:43:45 -0500, Vincent Rioux wrote:
>> Hello,
>>
>> I am trying to setup a software raid5 volume using Vinum on FreeBSD 5.2.
>> So far I did successfully complete these tasks :
>>
>>
>> - Created the config file :
>> drive d0 device /dev/ad10
>> drive d1 device /dev/ad11
>> (etc)
>
> These are disk devices, not disk partitions.  From the man page:
>
> DRIVE LAYOUT CONSIDERATIONS
>      vinum drives are currently BSD disk partitions.  They must be of type
>      vinum in order to avoid overwriting data used for other purposes.
> Use
>      disklabel -e to edit a partition type definition.  The following
> display
>      shows a typical partition layout as shown by disklabel(8):
>
>      8 partitions:
>      #        size   offset    fstype   [fsize bsize bps/cpg]
>        a:    81920   344064    4.2BSD        0     0     0   # (Cyl.
> 240*- 297*)
>        b:   262144    81920      swap                        # (Cyl.
> 57*- 240*)
>        c:  4226725        0    unused        0     0         # (Cyl.    0
> - 2955*)
>        e:    81920        0    4.2BSD        0     0     0   # (Cyl.    0
> - 57*)
>        f:  1900000   425984    4.2BSD        0     0     0   # (Cyl.
> 297*- 1626*)
>        g:  1900741  2325984     vinum        0     0     0   # (Cyl.
> 1626*- 2955*)
>
>      In this example, partition ``g'' may be used as a vinum partition.
> Par-
>      titions ``a'', ``e'' and ``f'' may be used as UFS file systems or ccd
>      partitions.  Partition ``b'' is a swap partition, and partition ``c''
>      represents the whole disk and should not be used for any other
> purpose.
>
>>   plex org raid5 433k
>
> The stripe size should be a multiple of the UFS block size, normally
> 16 kB.  Not doing so will not stop Vinum working, but it will impact
> on the performance.
>
>> Now the problem is that everytime I reboot the machine to test if
>> the volume automounts correctly (yes, start_vinum="YES" is in
>> rc.conf), all the configuration is lost and I must recreate and
>> reinitialize the volume (effectively losing all my data)...
>
> Well, no, just not finding it.  It's all still there.
>
>> Now what's weird, just after setting everything up and finishing
>> initializing the volume, I just can't seem to be able to dump the
>> config from the disks!
>
> This is a bug introduced in Vinum as the result of some recent changes
> in the bio layer.  Previously it would have refused to create the
> drives unless they were a partition of type "vinum".
>
>> Even if I do a saveconfig (which I shouldn't have to do since
>> config. updates are not disabled), there's nothing returned from
>> dumpconfig! :
>> vinum -> saveconfig
>> vinum -> dumpconfig
>
> Yes, if you don't have a configuration, you don't save anything.
>
>
> On Friday, 13 February 2004 at 17:04:45 +0100, Dag-Erling Smrgrav wrote:
>> "Vincent Rioux" <vrioux_at_centrenad.com> writes:
>>> I am trying to setup a software raid5 volume using Vinum on FreeBSD
>>> 5.2.
>>
>> Vinum does not play along with GEOM and should therefore be considered
>> broken in 5.x.
>
> No, that is not correct.  The only thing that doesn't work is swap on
> a Vinum volume.
>
>> Lukas Ertl is working on adapting Vinum to GEOM; see the
>> freebsd-geom list archives for more details.
>
> That is correct.
>
> Greg
> --
> See complete headers for address and phone numbers.
>
Received on Fri Feb 13 2004 - 18:35:25 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:43 UTC