FreeNAS Code

simplify label_disk (pointed out by gcooper)

Merge branch 'gitvolumes'

The time has come!
We still need to test upgrades from previous trunk, probably a few minor tweaks are to come

Conflicts:
gui/account/forms.py
gui/middleware/notifier.py
gui/services/forms.py
gui/storage/evilhack.py
gui/storage/urls.py
gui/templates/storage/dataset_confirm_delete.html
gui/templates/storage/dataset_edit.html
gui/templates/storage/datasets.html
gui/templates/storage/disk_detach.html
gui/templates/storage/geom_disk_replace.html
gui/templates/storage/permission.html
gui/templates/storage/scrub_confirm.html
gui/templates/storage/volumes.html
gui/templates/storage/zpool_status.html
gui/templates/storage/zvols.html
nanobsd/Files/usr/local/bin/custom-upssched-cmd
nanobsd/common

catch zfs snapshot errors

Unbreak afpd configuration reload.

Unfortunately Warner did something right and I broke the reloading the configuration because I didn't realize that notifier HUPed afpd with its PID.

We can get away with killall -1 afpd as it doesn't blow away existing share connections.

The hammer solution of using restart for everything but AD / LDAP in the services toggle form really needs to be fixed though because it emits a ton more noise than necessary and almost doubles the service setup and teardown time unnecessarily, which makes FreeNAS look braindead compared with other solutions.

At least this is a first step in the right direction...

use swapoff/swapon more wisely on zfs replace in case bad things happen

- Do a strip() before manupilating the public key;
- Remove an unneeded str() before calling save_pubkey.

Raise MiddlewareErrors on failed firmware upgrades and with basic service pack failures. Unfortunately errors aren't caught on service pack installs -- so caveat emptor!!!

cascade Task/Replication in case of dataset/volume/zvol delete

we cannot clone zvols, make the necessary changes to hide it for them

remove MountPoint from Replication, supposedly allow replication of zvols