Steve, We'd need a little more information about what went wrong. Was it a filesystem issue on /u01 of one compute node? If so (and you don't have backups), you could remove the node from the surviving cluster, and add it back in via the standard addNode.sh scripts. It's (more or less) documented in MOS note #1084360.1 - this includes extra information, such as if you had lost the compute node entirely and needed to reinstall the OS. If it was a storage server, there is a method to reimage the cell. All of that said, I would not go through the procedure myself without engaging someone that has done it before (or Oracle support). Andy Colvin Principal Consultant Enkitec andy.colvin@xxxxxxxxxxx http://blog.oracle-ninja.com On Jan 15, 2013, at 8:54 PM, Steve Harville wrote: > A client has corrupted the Linux file system which holds the software > Oracle installed on Exadata. There was no backup and the problem does not > have an easy fix. There are already some databases on the system and they > are OK. > Is it possible for customers to reinstall the software without disturbing > the databases? > Steve Harville > > http://SteveHarville.com > > > -- > //www.freelists.org/webpage/oracle-l > > -- //www.freelists.org/webpage/oracle-l