I have an lvm2 volume group composed of about 8 different pvs. Recently, most lvm commmands started displaying errors like below. Not that it is claiming duplicate PVs, but the duplicates are contained in /dev/nbd/*, and I'm not actually using any network block devices. I have NBD comiled into the kernel, and I'm about to compiling a kernel with nbd not compiled in and see if that helps, but it seems extremely odd that lvm would detect duplicate pvs on unused nbd devices.As I suspected, running a kernel with nbd compiled as a module and not loading it at startup avoids this probem. But this doesn't bring me any closer to an understanding of why lvm would see duplicate pv id's on nbd devices that I never used.
The effects of the problem (in addition to getting these warnings, are that any change to a lv(resizing, for example), will render that lv inactive or suspended (have seen both), until I do a magic incarnation that involves first "vgcfgrestore" and then "lvchange -a y" on the affected lv. Note that vgcfgrestore does not make the warnings go away, but it does allow the lvchange command to succeed.
Any thoughts? FWIW, I am using devfs, and I'm still curious about a reasonable migration strategy to move an lvm2 setup to non-devfs. Kernel 2.6.0-test11, lvm tools that come with debian sid(lvm version 2.00.05, library version 1.00.05-ioctl-cvs, driver version 4.0.0).
-Tupshin
-Tupshin
_______________________________________________ linux-lvm mailing list linux-lvm@sistina.com http://lists.sistina.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
