| 1 |
On Thu, 2006-04-06 at 15:50 -0500, Boyd Stephen Smith Jr. wrote: |
| 2 |
> On Thursday 06 April 2006 11:44, Fredrik Tolf <fredrik@×××××××××.com> wrote |
| 3 |
> about '[gentoo-user] LVM boot problem': |
| 4 |
> > I'm having a bit of a problem with LVM2 on Gentoo. See, I have a |
| 5 |
> > computer with a couple of hard drives in an LVM, and when it boots, I |
| 6 |
> > can see it loading the driver modules for the controller cards properly |
| 7 |
> > (I've added them to /etc/modules.autoload.d), but then when it comes to |
| 8 |
> > initializing the LVM, it complains that it cannot find all the PVs to |
| 9 |
> > initialize the VG. In fact, it complains about _all_ the PVs, as though |
| 10 |
> > it found none. |
| 11 |
> |
| 12 |
> Is / on LVM? Are you using an initrd / initramfs? |
| 13 |
|
| 14 |
Nope. Here's my fstab (minus pseudo-FS:s): |
| 15 |
/dev/hda2 / ext3 defaults,acl,user_xattr 0 1 |
| 16 |
/dev/hda1 none swap sw 0 0 |
| 17 |
/dev/vg1/newhome /home reiserfs defaults,acl,user_xattr 0 2 |
| 18 |
/dev/vg1/dlwsroot /srv/dlwsroot xfs defaults 0 3 |
| 19 |
/dev/vg1/site /usr/site xfs defaults 0 4 |
| 20 |
|
| 21 |
> Does your early userspace start a /dev daemon? |
| 22 |
|
| 23 |
If you mean udev, yes. |
| 24 |
|
| 25 |
> Does your early userspace have a different lvm.conf then your final userspace? |
| 26 |
|
| 27 |
Nope. |
| 28 |
|
| 29 |
> > Later, when the system is booted, I can log in as root and just run |
| 30 |
> > "vgchange -ay" to get the VG up. Note that I don't have to run pvscan |
| 31 |
> > first, which I'm assuming has to mean that it has somehow managed to |
| 32 |
> > find the PVs somewhere in the meantime. |
| 33 |
> |
| 34 |
> Are all of your pvs standard disks or disk partitions? |
| 35 |
|
| 36 |
Partitions. Here's the output of pvs: |
| 37 |
PV VG Fmt Attr PSize PFree |
| 38 |
/dev/hde1 vg1 lvm2 a- 111.79G 0 |
| 39 |
/dev/hdf1 vg1 lvm2 a- 111.79G 0 |
| 40 |
/dev/hdg1 vg1 lvm2 a- 111.79G 0 |
| 41 |
/dev/hdh1 vg1 lvm2 a- 111.79G 60.68G |
| 42 |
/dev/sda1 vg1 lvm2 a- 186.30G 186.30G |
| 43 |
/dev/sdb1 vg1 lvm2 a- 279.46G 0 |
| 44 |
/dev/sdc1 vg1 lvm2 a- 279.46G 0 |
| 45 |
|
| 46 |
|
| 47 |
> What modules are being loaded by the gentoo /etc/modules.autoload system? |
| 48 |
|
| 49 |
pcspkr |
| 50 |
pdc202xx_new |
| 51 |
sata_sil |
| 52 |
sd_mod |
| 53 |
rpcsec_gss_krb5 |
| 54 |
|
| 55 |
> What is your RC_VOLUME_ORDER in /etc/conf.d/rc? |
| 56 |
|
| 57 |
RC_VOLUME_ORDER="raid evms lvm dm" |
| 58 |
|
| 59 |
> Does your /lib/rcscripts/addons/lvm-start.sh |
| 60 |
> contain "/sbin/vgscan --mknodes --ignorelockingfailure >/dev/null"? |
| 61 |
|
| 62 |
Yep. |
| 63 |
|
| 64 |
> > This problem is quite annoying, since it means that I have to restart |
| 65 |
> > NFS and boot the system without the network cable plugged in (if it was, |
| 66 |
> > the NFS clients would try to reconnect and get handles to the wrong |
| 67 |
> > filesystem, which makes them mark the mount points as stale). |
| 68 |
> |
| 69 |
> You might investigate the mpt= /etc/exports option to only export |
| 70 |
> filesystems that have been successfully mounted, at least temporarily. |
| 71 |
|
| 72 |
Interesting. I'll definitely use that. For completeness, though, I |
| 73 |
should mention that looking at the manpage, it seems to be named "mp" |
| 74 |
rather than "mpt". |
| 75 |
|
| 76 |
Thanks! |
| 77 |
|
| 78 |
Fredrik Tolf |
| 79 |
|
| 80 |
|
| 81 |
-- |
| 82 |
gentoo-user@g.o mailing list |
Archives