2

I am trying out Solaris 11.3

Realize the option of -p when using beadm that i can actually create another boot environment on another pool.

root@Unicorn6:~# beadm create -p mypool solaris-1

root@Unicorn6:~# beadm list -a

BE/Dataset/Snapshot                              Flags Mountpoint Space   Policy Created

-------------------                              ----- ---------- -----   ------ -------

    solaris

       rpool/ROOT/solaris                            NR    /          3.35G   static 2018-02-07 02:21

       rpool/ROOT/solaris/var                        -     /var       429.15M static 2018-02-07 02:21

       rpool/ROOT/solaris/var@2018-04-06-01:38:54    -     -          20.0K   static 2018-04-06 09:38

       rpool/ROOT/solaris@2018-04-06-01:38:54        -     -          0       static 2018-04-06 09:38

    solaris-1

       mypool/ROOT/solaris-1                         -     -          3.44G   static 2018-04-06 09:38

       mypool/ROOT/solaris-1/var                     -     -          429.10M static 2018-04-06 09:39

       mypool/ROOT/solaris-1/var@2018-04-06-01:38:54 -     -          68.0K   static 2018-04-06 09:38

       mypool/ROOT/solaris-1@2018-04-06-01:38:54     -     -          83.38M  static 2018-04-06 09:38

I would have understood that snapshots are done the /rpool/ROOT/* datasets as shown above, then copied over to mypool

q1) What i dont understand is the need for the snapshots in mypool also ? The snapshots in rpool being copied over are consistent copies of the original BE. So why the need to snapshot those ROOT/solaris-1 datasets in mypool (ABE) again ?

Furthermore, it seems like this mypool/ROOT/solaris-1/var@2018-04-06-01:38:54 is created before the the actual dataset ?

mypool/ROOT/solaris-1/var                     -     -          429.10M static 2018-04-06 09:39

mypool/ROOT/solaris-1/var@2018-04-06-01:38:54 -     -          68.0K   static 2018-04-06 09:38

Can any kind souls shed some light on this ?

Regards, Noob

Noob
  • 363
  • 2
  • 6
  • 16

1 Answers1

1

They aren't snapshotted again. It's a zfs send/recv that's done, and that works by replicating a snapshot. The dataset is then created from that snapshot.

Dave Miner
  • 41
  • 1