Hey Fellas,

Before I continue please know I have a 3-2-1 solution, so I have 2 offsite backups, two entirely separate media onsite backups and a bunch of disks sitting around with backed up data. This is the main system but it's got so many copies "new" features are risk free. These are all regularly tested to.

I just finished manually adding 4 drives to 2 vdevs across a raidz2 array, and following the guides have re-mounted the pool successfully. Now I want to run the ZFS rewrite command as I understand it will re-stripe the array and be more optimal then leaving it alone space wise.

However the docs for the command are poor because its just come out. If you try to run it against the pool it just gives you a no such file or directory error and exits. Tabbing to return results gives you some options on selectable destinations but I dont think these are working properly yet.

So my question is... has anyone ran this successfully on 7.2.0 yet, and if so, what command did you use after zfs rewrite? How did it work, what was your experience etc.

Cheers.

You can run it and it works. Note that if you have snapshots or use incremental send/receive, that can cause more space to be used and/or sending all the data again.

Also note that you must specify the path, not the pool/dataset, e.g.:

zfs rewrite -r /mnt/tank

You can get an kind of ETA with the -v argument but need to keep the web console opened. The -P argument shall come within 2.4.

Thank fellas, that's done it. We're in progress for a re-write.

I'll delete my snapshot to reclaim the space officially once it completes and I verify basic data integrity.