Hello,
we have a FreeNAS backup 'BU1' that probably will need to have its pool rebuilt because of some h/w issues.
Currently there is an incremental replication established from BU1 to remote server BU2.
However, probably we'd need to rebuild BU1 pool, and because the bandwidth with BU2 wouldn't allow a fast copy
I'm building a new (temporary) local server BU3 (otherwise full recovery from BU2 would take a week)
So the idea is:
- full (fast!) replication BU1 to BU3
- stop incremental to BU2
- rebuild BU1 pool (as it was before)
- restore BU1 from BU3
- restart incremental BU1 -> BU2
(- decommission BU3)
Now the question is: in such scenario, can the incremental replication continue from where it was left,
or would BU2 detect that BU1 pool has been rebuilt and basically the replication has to start from scratch?
E.g. is there some source dataset/zpool UUID in play to be preserved to avoid that?
Thanks
we have a FreeNAS backup 'BU1' that probably will need to have its pool rebuilt because of some h/w issues.
Currently there is an incremental replication established from BU1 to remote server BU2.
However, probably we'd need to rebuild BU1 pool, and because the bandwidth with BU2 wouldn't allow a fast copy
I'm building a new (temporary) local server BU3 (otherwise full recovery from BU2 would take a week)
So the idea is:
- full (fast!) replication BU1 to BU3
- stop incremental to BU2
- rebuild BU1 pool (as it was before)
- restore BU1 from BU3
- restart incremental BU1 -> BU2
(- decommission BU3)
Now the question is: in such scenario, can the incremental replication continue from where it was left,
or would BU2 detect that BU1 pool has been rebuilt and basically the replication has to start from scratch?
E.g. is there some source dataset/zpool UUID in play to be preserved to avoid that?
Thanks