I'm running backuppc 3.2.1 on Linux Mint 11, backing up a machine (A) which
has SMB (as opposed to rsync/tar/...) as the only available backup
mechanism.
An initial full backup of A is fine, but a subsequent incremental backup
does not backup files which were added to A since the full backup.
It
I'm running backuppc 3.2.1 on Linux Mint 11, backing up a machine (A)
which
has SMB (as opposed to rsync/tar/...) as the only available backup
mechanism.
This is probably something worth reconsidering.
An initial full backup of A is fine, but a subsequent incremental backup
does not
On Fri, Nov 2, 2012 at 10:45 AM, Jim Stark jstarkjhea...@gmail.com wrote:
backuppc documentation says: For SMB and tar, BackupPC uses the
modification time (mtime) to determine which files have changed since the
last lower-level backup. That means SMB and tar incrementals are not able to
Thanks for the prompt reply.
Machine A is an old Buffalo TeraStation. Ironically, it is running Linux
likely even has rsync, but as a practical matter, SMB is the only available
access mechanism.
1) You can touch the files, or copy them in such a way that their times
change
No shell access to
Jim Stark jstarkjhea...@gmail.com schrieb:
Thanks for the prompt reply.
Machine A is an old Buffalo TeraStation. Ironically, it is running
Linux
likely even has rsync, but as a practical matter, SMB is the only
available
access mechanism.
1) You can touch the files, or copy them in such a way
On Fri, Nov 2, 2012 at 11:37 AM, Jim Stark jstarkjhea...@gmail.com wrote:
2) You can simply run full backups
Performance hit? I understand pooling will avoid creating multiple copies,
but cost in backup time?
You need to do fulls at least once a week or so to keep the tree
structure sane -
Mount the buffalo with smbfs or similar into a unix/linux machine and let
touch run.
Yes, thanks. I could do that.
On Fri, Nov 2, 2012 at 11:11 AM, Arnold Krille arn...@arnoldarts.de wrote:
Jim Stark jstarkjhea...@gmail.com schrieb:
Thanks for the prompt reply.
Machine A is an old
My concern is that to verify that a file can be pooled, it first has to be
brought over from machine A (true?).
Thus, even if it is ultimately pooled only costs 1 hard link in the
backup for the host, there is all of the full backup overhead on machine A
and all of the network traffic to get it
On Fri, Nov 2, 2012 at 2:02 PM, Jim Stark jstarkjhea...@gmail.com wrote:
My concern is that to verify that a file can be pooled, it first has to be
brought over from machine A (true?).
Thus, even if it is ultimately pooled only costs 1 hard link in the backup
for the host, there is all of