When importing a pool the enclosure sysfs path stored in the label should be updated and validated.
The current code only updates the stored path when the enclosure sysfs path can be resolved for the given vdev. If the block device cannot be opened by any known path and is UNAVAIL it is currently not updated.
What we'd like to do is use the enclosure sysfs path stored in the label to reconstruct and verify the path. even when the block device is unavailable. This could be done by storing the unique enclosure ID in the vdev label and using it to locate the correct enclosure then appending the previous slot ID.
If the above fails then the enclosure sysfs path should be removed from the vdev label.
Lastly we should update the ZED to set the fault led for any vdevs which are unavailable at import time and which have verified enclosure sysfs paths.
Import a pool with an UNAVAIL disk.