I'm searching rather urgent help for the following issue:

I have a groupwise cluster that allways worked fine.
Last week we replaced the hardware of 1 node, since then i have the
following issue:

NODE A and NODE B have the same 3 luns with each their partition.(SBD ,
NODE A works fine a has at this moment the full load.
NODE B sees the 3 luns, but considers 1 of them as empty. NODE B doesn't
detect on 1 lun the partition & pool.

I tried already some things :
- reinstalled node from scratch
- checked driver for upgrade
- took a copy of the luns in testenvironment and did a poolrebuild (on
node that detects it)
- deletect and rebuild SDB partition
- removed SERVCFG.* files in c:nwserver

I'm completely out of ideas.
Is there someone who recognise this kind of problem ?

thanks in advance for you feedback.