On Jan 25, 2014, at 4:12 PM, Adam Williamson <awilliam(a)redhat.com> wrote:
* Do an offline update that includes Foo v2.0
* Boot the updated system, run Foo, it migrates its configuration to
some new scheme
* Realize there was something wrong with the update, roll it back
* Run Foo again, find it doesn't work because it's been migrated to the
new config scheme which the old version of Foo doesn't work with
I would grumble, but a configuration file being updated and made incompatible with the
prior version would be tolerated. Ideally the application makes an unmodified copy. If it
doesn't, new school restore with --reflink from snapshot, regular cp if using LVM
thinp snapshots, and old school just restore the file from a conventional backup. Not such
a big deal.
If it's something far less throw away than configuration files being changed, it's
a bit more complicated how badly and quickly the conversation degrades. But I can hardly
recall a recent example of this happening. It's just not that common in my
experience.
Chris Murphy