vinum drive referenced / disklabel inconsistency

[Clark C. Evans]

Perhaps I was being too clever trying to setup a swap partition,
beacuse If I change my label so that vinum  part is idential to
the "unused" part (but assigned to a:) it works.   *puzzles*

Best,

Clark

On Sun, Jun 29, 2003 at 08:12:06PM +0000, Clark C. Evans wrote:
| I am trying to setup vinum on a box using 4.8 RELENG_4 (as of about a
| week ago snapshot).  This box was running 4.6 /w vinum on same hard 
| drives for the last 4 months wonderfully... but since it is my 
| current 'scratch/backup' box, I just reinstalled with -STABLE.
| 
| # uname -a
|   FreeBSD polya.axista.com 4.8-STABLE FreeBSD 4.8-STABLE #22: Tue Jun 24
|   17:01:07 EDT 2003     cce at cantor.local:/usr/obj/usr/src/sys/GENERIC i386
| 
| Anyway, my problems seem to be coming from a disklabel issue.  Following
| the instructions in the handbook (for disk addition), I go through the
| following steps:
| 
|    # dd if=/dev/zero of=/dev/ad2 bs=1k count=1
|    # fdisk -BI ad2
|    # disklabel -B -w -r ad2s1 auto
| 
| And at this point, I use the following to extract the current
| label, and then I edit it with vi, and then update it:
|   
|    # disklabel ad2s1 > label
|    # vim label
|    # disklabel -R -r ad2s1 label
| 
| Following is the tail end of the label (I don't edit anything else)
| where I add 'swap' and 'vinum' partitions:
| 
|   a: 4193280 0 swap
|   c: 156301425        0    unused        0    0 
|   e: 152107200 4193280 vinum
| 
| The results, when "restoring" the label, however, are inconsistent, with
| 
|   # disklabel ad2s1
|   # disklabel ad2
|   # disklabel -r ad2s1
|   # disklabel -r ad2
|   # disklabel /dev/ad2s1
|   # disklabel -r /dev/ad2s1
| 
| the results are expected (and identical):
| 
|   <snip/>
| 
|   8 partitions:
|   #        size   offset    fstype   [fsize bsize bps/cpg]
|     a:  4193280        0      swap                # (Cyl.    0 - 4159)
|     c: 156301425        0    unused        0    0 # (Cyl.    0 - 155060*)
|     e: 152107200  4193280     vinum               # (Cyl. 4160 - 155059)
| 
| However, for the other two permutations... the results are different: 
| 
|   # disklabel /dev/ad2
| 
|   <snip/>
| 
|   8 partitions:
|   #        size   offset    fstype   [fsize bsize bps/cpg]
|  c: 156301488        0    unused        0     0   # (Cyl.    0 - 155060)
|   
|   # disklabel -r /dev/ad2
|   disklabel: bad pack magic number (label is damaged, or pack is unlabeled)
| 
| At this point, I use the following vinum configuration file to
| setup vinum "vinum create -f vinum.conf"
| 
|     drive d2 device /dev/ad2s1e
|     volume tmp setupstate
|      plex name tmp.2 org concat
|       sd length 2000m drive d2
| 
| And this produces...
| 
|   1 drives:
|   D d2                State: referenced       Device  Avail: 0/0 MB
|   
|   1 volumes:
|   V tmp             State: down     Plexes:       1 Size:       2000 MB
|   
|   1 plexes:
|   P tmp.2         C State: faulty   Subdisks:     1 Size:       2000 MB
|   
|   1 subdisks:
|   S tmp.2.s0        State: crashed  PO:        0  B Size:       2000 MB
| 
| 
| And of course, this is where I get stuck...  With 4.6 I didn't have any
| of these problems (from the bad pack magic number to the missing device
| in the vinum state).   Anyway, following is the full output from the
| disklabel.
| 
| # disklabel -r /dev/ad2s1
|   # /dev/ad2s1:
|   type: unknown
|   disk: amnesiac
|   label:
|   flags:
|   bytes/sector: 512
|   sectors/track: 63
|   tracks/cylinder: 16
|   sectors/cylinder: 1008
|   cylinders: 155060
|   sectors/unit: 156301425
|   rpm: 3600
|   interleave: 1
|   trackskew: 0
|   cylinderskew: 0
|   headswitch: 0           # milliseconds
|   track-to-track seek: 0  # milliseconds
|   drivedata: 0
|   
|   8 partitions:
|   #        size   offset    fstype   [fsize bsize bps/cpg]
|     a:  4193280        0      swap                  # (Cyl.    0 - 4159)
|     c: 156301425        0    unused        0     0  # (Cyl.    0 - 155060*)
|     e: 152107200  4193280     vinum                 # (Cyl. 4160 - 155059)
| 
| Thank you for any suggestions you may have,
| 
| Clark
| _______________________________________________
| freebsd-stable at freebsd.org mailing list
| http://lists.freebsd.org/mailman/listinfo/freebsd-stable
| To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"