W.Kenworthy wrote:LVM the gentoo way works fine: I highly recommend for any system needing multiple partitions
http://www.gentoo.org/doc/en/lvm2.xml
I would agree and this is a good document to follow. But unfortunately, it does not cover the udev case.
following the instructions in the above document,
I set up a raid 1 system on md0 (partition type fd)
I then created all of the elements for a volume group
mounting and copying files works fine
I then discover I need to run udev
I convert according to documents other people have given me
it mostly works okay except that the the raid array is no longer being detected. It appears that the raid code is invoked before the modules or the devices for the external ide array are available.
So that's the path I'm currently investigating.
bah, humbug
---eric
As a workaround while you do so, I would suggest using the devfs tarball (which should contain the devices you previously had before you switched to udev)-- this is set in /etc/conf.d/rc :
RC_NET_STRICT_CHECKING="no"
# Use this variable to control the /dev management behavior. # auto - let the scripts figure out what's best at boot # devfs - use devfs (requires sys-fs/devfsd) # udev - use udev (requires sys-fs/udev) # static - let the user manage static nodes
RC_DEVICES="auto"
# UDEV OPTION: # Set to "yes" if you want to save /dev to a tarball on shutdown # and restore it on startup. This is useful if you have a lot of # custom device nodes that udev does not handle/know about.
RC_DEVICE_TARBALL="yes"
You might also just switch back to devfs (compile both udev and devfs into your kernel, and then use the bootloader config to choose which one to use; I think the setting is something along the lines of nodevfs or something, but it is documented).
I'm not sure, but it seems to me that the issue is something with RAID rather than with LVM; are your RAID control modules (other than dev-mapper, I mean the modules for the card or mobo controller or software RAID) compiled directly into the kernel, or as modules?
Holly -- [email protected] mailing list
