Shared Volume Group not deactivating with vgchange -a n <vg>

[Matthew Murdock]

2 node Service Guard Cluster with a shared Volume group/logical volume, with Oracle 11g RAC. De-actiavtion of the Shared Logical volume group use to work (vgchange -a n <vg>). 

Now after Oracle is shutdown with "crsctl stop has", and all the oracle processes are dead, (ps -ef shows no oracle user on the box) the vg cannot be deactivated (vgchange -a n <vg>). It complains about a lock cannot be stopped - /etc/lvmconf/pv_lock is still open (was up till 430 am trying to get it to work) or something of the sorts.....Is there some way to find what process has the shared logical volume still open?

 If the node is shutdown it panics the system because the os will not release the lock on the shared VG. IN MY CASE, a LV RESYNC of the same VG/LV we are trying to close, on the opposite node was occuring at the same time. Will this prevent the Shared volume group/LV on the Opposite node from deactivating or would I have something else holding the LV lock?

We are running the Sept.15 of the Quality Pack patches on the nodes.