ZFS should provide option to automatically promote hot spares to replacements
autoreplace zpool option introduced in PSARC/2007/197 allows for hot swapping of faulted component devices with new devices located at the same physical device path without issuing system commands. This unfortunately does not work with SAS, iSCSI, and other storage systems where the device location is not determined by the physical layer configuration (the case suggests extending this with information from libtopo).
One alternative is to designate hot spare devices. They are automatically added as substitute pool members by FMA when a device is found faulty. However, they remain on the hot spare list until the faulted device is @zpool detach@ed, at which point they're promoted to a replacement and removed from the hot spare list.
Adding an option to automatically promote them to replacements would be useful for an application where they're used in the same vein as