We're planning our replication strategy for migrating from one datacenter to another, and the lack of an easy re-ip for Linux is something I keep coming back to. I'd rather not have a script that runs at boot time and tries to figure out where the VM is based on the pingability of a local gateway or some other clever means. Instead, I would like to say that whenever the replica is written, run a script on the VM which will replace the /etc/network/interfaces (in the case of an Ubuntu VM) or /etc/sysconfig/network-scripts/enoXXXXXXXXX (CentOS) file with an alternate version already on the VM. Each time a new replica delta is taken, it will re-run this script so that the replicated VM always has its IP configuration for the destination site, without introducing a script that runs at boot time that could potentially cause issues with the VM at the original site.
Is that how the pre-freeze/post-thaw script rules work? Will it run the script at the time of capture (pre-freeze) or at the time the replica is powered on (post-thaw)? If that's the case, isn't it true that it could be used for more than just quiescence? Do both pre-freeze and post-thaw only touch the replica, or does pre-freeze refer to the original VM, and post-thaw the replica?
Thanks in advance!