vldb repair not recreating the ~DFSINFO.8-P for cluster enabled volumes.

  • 7012206
  • 25-Apr-2013
  • 25-Apr-2013


Novell Open Enterprise Server 2 (OES 2) Linux Support Pack 3
Novell Open Enterprise Server 11 (OES 11) Linux
Novell Open Enterprise Server 11 (OES 11) Linux Support Pack 1
Novell Cluster Services (NCS)
Novell Distributed Files Services (DFS)


If the ~DFSINFO.8-P file is overwritten by one of an other volume, is deleted or got corrupt, than junctions are no longer functional and new junctions will point to a wrong volume.
The ~DFSINFO.8-P file can be recreated using the steps outlined in the documentation for OES2SP3 and OES11 by running a vldb repair or vldb add.
This procedure however will not work for cluster (NCS) enabled volumes.


A script, which will be distributed on demand trough Novell Technical Services, is in the making that can be used until vldb is adjusted so it can also handle cluster volumes.

However, a current workaround is to go trough these steps:

1. run a vldb repair (which should create the new ~DFSINFO.8-P, but does not for a cluster volume)
2. Copy a non corrupt ~DFSINFO.8-P from a working volume as base for the new file for the affected cluster volume.
3. Copy the GUID from the eDirectory volume object.
4. Past the GUID correctly into the ~DFSINFO.8-P using a HEX editor of your choice (like for instance UltraEdit).
Be Aware: The entry should be added in HEX, preserving the preceding FF FF 00 00 10 00 00 00 so it should look like: 
FF FF 00 00 10 00 00 00 xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx xx
(where the xx form the volumes GUID).
5. Copy the newly adjusted ~DFSINFO.8-P to root of the affected volume.
6. Verify the owner and rights so they match: rrr root:root
7. Delete and recreate the junction toward this volume (as per documentation).


vldb repair checks the host server of the volume, if this is set to a cluster virtual server, the command is unable to determine the state of the server and skips this volume's ~DFSINFO.8-P.


Reported to Engineering