../bin/$BFARCH/OprCmd.pl -iDebug,User,Farm -noprserv0x farminfo XXx
 you will see one (or more) nodes that are
still in 'Started', i.e.

 IP             -        Port   -      Elf ID   -         Status
...
10.100.2.117    -        1487   -        034    -        Finished
10.100.2.123    -        1353   -        051    -        Started
10.100.2.111    -        1560   -        019    -        Finished

sometimes the copy will fail while the others are still copying,
so you want to wait until all the nodes are done except the ones
that failed... or you can get the elf number from the email... 

check the current state of the NP, i.e.

pulliam@bbr-user:ER3:prod/workdir> ../bin/Linux24RH72_i386_gcc2953/OprCmd.pl -iFarm,User,Debug -n10.100.2.123 -p 1353 npcurstate
connected
Message [OprMessage=HASH(0x86cec78)]:
From       : OprFSMFwkModule:NP
To         : OprMpxServer
Content    : AnswerGetCurrentState
Attributes : FatalError

the Fatal Error means it is the right one... 

then set the state:

pulliam@bbr-user:ER3:prod/workdir> ../bin/Linux24RH72_i386_gcc2953/OprCmd.pl -iFarm,User,Debug -n10.100.2.123 -p 1353 npsetstate CopyFilesToServerRoo
connected
Message [OprMessage=HASH(0x86cef14)]:
From       : OprFSMFwkModule:NP
To         : OprMpxServer
Content    : AnswerSetCurrentState
Attributes : OK forced transition to CopyFilesToServerRoo

and check... 

pulliam@bbr-user:ER3:prod/workdir> ../bin/Linux24RH72_i386_gcc2953/OprCmd.pl -iFarm,User,Debug -n10.100.2.123 -p 1353 npcurstate          
  connected
Message [OprMessage=HASH(0x86cec78)]:
From       : OprFSMFwkModule:NP
To         : OprMpxServer
Content    : AnswerGetCurrentState
Attributes : CopyFilesToServerRoo

and then the run should finish on it's own...