On 2013-11-14 01:41, jason matthews wrote:
zfs list is reporting the number you are interested in. zpool list essentially reports raw data which isn't particularly useful for capacity planning.
...And some of the discrepancies (between logically available and physically allocated data counts) are due to: 1) RAIDZ Parity - not-yet-allocated parity sectors don't count in zpool, but do count as (smaller) available logical data in zfs list. 2) Reservations, such as for volumes and for default volume snapshot reservations (allowing to rewrite all of the volume) - this space is not available to other datasets (including volume root), but some or none or all of this may be allocated as physical data. 3) 1/64 of the pool size is AFAIK reserved for internal housekeeping. In one of our cases this sums up to a largish difference: # df -h /pond Filesystem size used avail capacity Mounted on pond 8.0T 34K 174G 1% /pond # zfs list pond NAME USED AVAIL REFER MOUNTPOINT pond 7.85T 174G 34.3K /pond # zpool list pond NAME SIZE USED AVAIL CAP HEALTH ALTROOT pond 10.2T 9.37T 845G 91% ONLINE - HTH, //Jim PS: Don't start new threads by replying to older unrelated emails ;) _______________________________________________ OpenIndiana-discuss mailing list [email protected] http://openindiana.org/mailman/listinfo/openindiana-discuss
