Why rpool/swap volume uses a USEDREFRESERV less than REFRESERV ?
Hello,
I've got to know a little more about ZFS volumes but, in particular, I still don't understand why rpool/swap doesn't realize all of its reservation as ordinary volumes.
Furthermore, rpool/swap is not a sparse volume such as rpool/dump which has REFRESERV = none.
What motivates my questioning is that this may have consequences (as happened to a coworker of mine) on (old) systems using ZFS without enough AVAIL in rpool for an eventual snapshot of rpool/swap.
Of course I can't envision any special reason for snapshotting rpool/swap but it can happen as part of a disaster recovery procedure described in the official documentation.
I've got to know a little more about ZFS volumes but, in particular, I still don't understand why rpool/swap doesn't realize all of its reservation as ordinary volumes.
Furthermore, rpool/swap is not a sparse volume such as rpool/dump which has REFRESERV = none.
What motivates my questioning is that this may have consequences (as happened to a coworker of mine) on (old) systems using ZFS without enough AVAIL in rpool for an eventual snapshot of rpool/swap.
Of course I can't envision any special reason for snapshotting rpool/swap but it can happen as part of a disaster recovery procedure described in the official documentation.
0