Hello All,
I have a few questions about disaster recovery with Netbackup.
Some Background:
I have worked with ufsdump and Legato Networker quite a bit, and have a
*little* bit of experience with Netbackup
Platform:
Sun Low End (E220R), Mid Range (E4500,E6500) to High end (E10K)
Considerations:
In a DR situation, we may not get the same exact platform to restore to so
restoring the root volume is not an option unless we want to play games with
drvconfig, /devices, path_to_inst, etc.. (eg. restore E420R to E5500, PCI based
to SBUS or vise versa)
Therefore we would need to install the OS, then restore the data partitions
with Netbackup and play the games of "what did we forget from the root
partition" and restore those files selectively. i.e. jni configuration from
/kernel/drv, /etc/hosts, /sbin/rc scripts, etc..
We assume that a 747 hit our data center, therefore all we have to start with
is a bunch of tapes from the vault (offsite storage) and a bunch of new Sun
machines...
Questions:
Assuming that the Netbackup Master server was backed up with Netbackup and only
a single stream (not multiplexed or anything like that), can we expect to send
a floppy offsite with gnu tar on it and restore the Netbackup Server?
Would it be a better idea to send a tape offsite with a ufsdump of the
Netbackup server and use that?
I have always approached bare metal recover as a complete restore from tape
(ufsdump) but since we can't be sure that we get the same platform, I think
this isn't really an option.
Comments please ...
TIA
--Buddy
|