Hi all.
Does anyone know why offhand bprestore requires a specific end time
(down to the seconds) when performing a True Image Restore? The
manual page state's its to uniquely identify the image to be restored,
but I'm not sure that I fully understand this. Or maybe I'm trying
to use bprestore for something it wasn't meant to do.
Basically I would like to able to use bprestore to perform a full
filesystem restore in a DR situation. The scenario would go something
like,
- boot single-user from a recovery volume (min os + nb client)
- startup inetd to allow nb communications
- initialize volumes to recovered
- mount volumes to be recovered on an alternate path
- use bprestore to initiate the restore process
Initially I though I could use a command such as,
# bprestore -H -T -L /tmp/restore.log -R /tmp/restore.map \
-C client.domain-name -S server.domain-name /
and have it perform a TIR restore using the most recent full and
cumulative backups. But issuing this command without a specific end
date that matches the last cumulative backup, the backup just exits
saying that nothing matched the requested selection.
The documentation suggests using the 'bplist -l' command to date and time
down to the seconds, but this is a bit confusing to me. Suppose I issue the
command,
bplist -b -l -Listseconds /
which displays the backup times for the directory /, which according to the
docs should be used as the anchor time (-e). But does this imply that the
restore operation will only restore files from that backup? Or if the
backup time for / is prior to the last backup time for /var or /usr, does that
imply that the most recent files for /var and /usr would not be restored?
Hoping someone can shed some light on this.
Thanks,
John
--
John Meyers
Computing Services
Wright State University
E-mail: john.meyers AT wright DOT edu
|