Old FreeBSD7.0 filesystems in FreeBSD8.0

Q

qcure

Hi guys,


Long time reader and first time poster here. Yesterday reinstalled my box with FreeBSD8.0, I had a ZFS on the old FreeBSD7 and I needed to recreate that (wanted to make it GPT with labels) so I backed up my data to other drives, that were sliced and partitioned the old fasion way via sysinstall. The FreeBSD8 install was made this way -> http://m8d.de/news/freebsd-on-gpt.php, since sysinstall refused to slice and partition the system disk with some weird errors ( never had problems with 6.4 when I done it).

The FBSD got installed and all is fine, but now that I recreated the ZFS I plugged the disks with my data, freebsd picks them up but in /dev/ I noticed that the disk appeared as /dev/ad6p1.../dev/ad6p2 and not the old fasion way /dev/ad6s1c or something. Trying to mount the /dev/ad6p1, fails with "invalid arguments" after the mount command. So not knowing what was going on I booted a older version of PC-BSD ( didn't have freebsd near me ) and the disks showed the proper way and mounted fine, the data was there and all was fine.

So my question is what do I have to do in order to mount the drives in freebsd 8, I noticed that in dmesg of the PC-BSD there were some messages GEOM_LABEL and etc, that were missing from the FreeBSD 8. I tried to
Code: 
[root@razor /boot/kernel]# kldload /boot/kernel/geom_label.ko
kldload: can't load /boot/kernel/geom_label.ko: File exists

So any help will be thankful to get my data back on my raidz. Thanks in advance.
 
If I understood you correctly, then I suppose an upgrade from sources is the right choice, compile world and kernel. Just make sure you merge your configs.
 
Can't do that since the old BSD is gone. All I need to know is how can I make the old slices/partitons properly visible in clean FreeBSD8 installation.
 
ad6p1 means that this disk is GPT partitioned, not the "old" slice way.
What does 'ls /dev/ad* /dev/da*' show?

Are you sure you are looking at the correct disk ?
[dmesg should show what disk is what /dev/ node]

[root@razor /boot/kernel]# kldload /boot/kernel/geom_label.ko
kldload: can't load /boot/kernel/geom_label.ko: File exists
Click to expand...

means the module is already loaded
[ls /dev/label ?]
 
Code: 
ls -l /dev/ad6*
crw-r-----  1 root  operator    0,  90 Jan 30 23:15 /dev/ad6
crw-r-----  1 root  operator    0,  95 Jan 30 23:30 /dev/ad6p1
crw-r-----  1 root  operator    0,  96 Jan 30 23:30 /dev/ad6p2

 ls -l /dev/da0*
crw-r-----  1 root  operator    0, 133 Jan 31 12:52 /dev/da0
crw-r-----  1 root  operator    0, 134 Jan 31 12:52 /dev/da0p1
crw-r-----  1 root  operator    0, 135 Jan 31 12:52 /dev/da0p2

gpart show

=>       34  390721901  ad6  GPT  (186G)
         34          6       - free -  (3.0K)
         40     409600    1  efi  (200M)
     409640  390050144    2  !48465300-0000-11aa-aa11-00306543ecac  (186G)
  390459784     262151       - free -  (128M)

=>       34  390719788  da0  GPT  (186G)
         34          6       - free -  (3.0K)
         40     409600    1  efi  (200M)
     409640  390048032    2  !48465300-0000-11aa-aa11-00306543ecac  (186G)
  390457672     262150       - free -  (128M)

Those are the 2 disks, that I can't mount. The da0 disk in older version of FreeBSD, shows the proper way with the slices and I can mount with no problems. On older versions of FreeBSD when I plug in the da0 disk I see, GEOM_LABEL message in /var/log/messages. I don't see those messages in FreeBSD8...

No such thing as /dev/label.
 
3 other ko's you may want to add to /boot/loader.conf (or manually)
Code: 
geom_bsd
geom_mbr
ehci
(_load="YES" ) iirc (each of them).
(Might pertain to the problem.) Might not...
(I only have about 2 minutes total to think up an
answer to, and even read the thread) as of right now.
 
I would say I love you, but i'm not that kind of guy :D. Thanks alot, that fixed it. On the old FreeBSD, I didn't saw those lines in the /boot/loader.conf and it still worked...strange.
 
You must log in or register to reply here.
Back
Top