[Spacewalk-list] Back to the same problem after snapshot restore

[m.roth2006 at rcn.com]

I thought we'd resolved this, but neither I nor the other admin I work with can figure this.

CentOS 5.2, x86_64 client, ia32 server, VMware ESX guests, both of them. Spacewalk 0.4

I restore the client from a snapshot I took before doing my original upgrade, so it's built from one of our templates. 

We have:
   restarted osad
   verified packages
   (try #1 to upgrade)
   deleted the system from the server
   rm /etc/sysconfig/rhn/systemid
   register the client
   yum clean all
   yum check-update
   (try again)
Somewhere in there, I ran rhn-profile-sync
 
I'm down to trying to upgrade one single package, and it fails, with 
Error while executing packages action: Transaction Check Error: file /usr/share/man/man8/dmraid.8.gz from install of dmraid-1.0.0.rc13-15.el5_2.1 conflicts with file from package dmraid-1.0.0.rc13-9.el5 [[6]]

This makes *utterly* no sense to me, since it's an upgrade to the latter. 

Oh, and the other admin tried to remove a single package, and right above where it said this was scheduled, it failed; in failed systems, it reports:
invalid action 317 for server 1000010042 


Ok, folks, what have we missed?

   mark

PS: spacewalk enhancement request: "\n" after each failed package.