Backing up SCO Unix/Unixware with BE 9.1- problems

[neoadmin]

I'm currently using Backup Exec v9.1 for Windows Servers to backup my entire network of Windows and Unix based machines. All Windows boxes appear to be backing up fine, no problems there at all, but with any Unix system (the ones we have being SCO Unix or Unixware) we are having problems with the job completing the backup of the systems using the Unix Agent but being flagged in BE9.1 as 'Completed with exceptions'.

The problem appears to be that the several files on all the unix systems are being skipped, and they are not the same filenames, just seem to be getting a lot of:

'File XXXX was not found, or could not be accessed.'

I've tried setting each of the Unix Agent's extra options such as 'include_remote' and 'no_nfs_locking' as stated in the Veritas support article

http://seer.support.veritas.com/docs/231634.htm,

but am still experiencing the problem.

Any BE guru's have any ideas on this? would be much appreciated.

Thanks

 

[neoadmin]

*bump* - nobody have any ideas on this at all still????? - still having the problem.

 

[dooza]

first: be sure the file you want to backup is being there.
second:be sure the user can access the file which you want to backup.
example:if you want to backup the file "test.conf",the backup server access the client use the account "user".so the account "user" must has permission to read the file "test.conf",you can change it as root via "chmod 644 test.conf“?

try it!

 

[neoadmin]

first: the files being backed up *are there*
second: I'm pretty sure the BE server can access the files to back them up, as most of them have the right privelidges set (as with your example) and still have the error message:

"File xxx/xxx was not found, or could not be accessed."

It seems though that most of the files its having problems with are 'pipe' files (have p attribute) - any more ideas?

thanks