why not keep going?

[Robert.Weber@Colorado.EDU]

I am using V  0.6.1.
	I am setting up a large scale backup and I'd like to use rdiff
backup.  I have noticed that if the rdiff fails for some reason(the file
contents change while it is running) it kills the ENTIRE backup.  I've got
an active filesystem and I keep getting

__main__.RdiffException: ['rdiff', 'signature', '/backup/animas/projects/hdb/staff/caseyh/oracle_script/TRIGGERS/CVS/Entries'] exited with non-zero value -1

and it kills the backup.  Why not either 

a) copy files to /tmp before doing the sig

b) move on to the next file?

Also, you have code to identify some special files, but apparently not
"door" files.  these should not be dumped, and becuase of above, it
prevents / backups on some systems(door files go in /etc on solaris).

(animas)[10:56am]:~ > ls -l /var/run/syslog_door 
Drw-r--r--    1 root     root            0 Apr 17 18:19

If you could identify them that would be great, however if b) from above
was implemented then it wouldn't be an issue.


						Robert Weber
						University of Colorado