Comment 10 for bug 332270

The problem seems to come from the watch option added in 60-persistent-storage.rules.

1. Booting with break=premount drops me in busybox without problems. With break=mount the disk activity already starts.
2. Booting with break=premount, then starting udevd and calling udevadm trigger --action=change starts the disk activity
3. Same as 2, but first removing the watch options from 60-persistent-storage.rules and 65-dmsetup.rules does not trigger the disk activity. However, the disk activity now starts after / is mounted (/ is on lvm).
4. Same as 3, but also removing the watch option from the same 2 rules files in / makes the boot successful.
5. Setting udev's loglevel to debug, adding the watch option in 65-dmsetup.rules and connecting an external usb drive does not cause any problems.
6. Same as 5, but also adding the watch option to 60-persistent-storage.rules results in the high disk activity. syslog is quickly filled with messages.
7. After removing the watch option from 60-persistent-storage.rules again, I can connect the usb drive without problems again.

Attached is the syslog part from step 6, between connecting the usb drive and disconnecting it again.
The most surprising lines for me are:
Feb 21 13:29:20 jaunty udevd-event[16733]: device will be watched for changes
Feb 21 13:29:20 jaunty udevd-event[16733]: 'watershed sh -c '/sbin/lvm vgscan; /sbin/lvm vgchange -a y''

Why should vgscan start again after the watch is added?

The usb drive has 4 partitions. 2 are ext3, 1 is swap and 1 is LVM:
$ sudo fdisk -l /dev/sdc

Disk /dev/sdc: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000e7a91

   Device Boot Start End Blocks Id System
/dev/sdc1 * 1 6119 49150836 83 Linux
/dev/sdc2 6120 12238 49150867+ 83 Linux
/dev/sdc3 12239 18357 49150867+ 8e Linux LVM
/dev/sdc4 18358 19457 8835750 82 Linux swap / Solaris