surfrock66
Dabbler
- Joined
- Apr 2, 2013
- Messages
- 36
I needed to pull a file out of a snapshot, but I did something wrong and seem to have pretty seriously messed up my setup. Fortunately this is a backup location for live data running on other servers so nothing was really lost, but I'm confused what I did and how to fix where I'm at.
I needed to pull a file I overwrote from a snapshot. I found the snapshot, in the gui I cloned it to a new dataset. I went in via ssh, copied what I needed out, then tried to delete the dataset, and the option was greyed out. I googled it, they said I had to promote the new dataset using "zfs promote" which I did, then the result said to do a zfs destroy of the dataset, which prompted for a -r and -R. I did that, and it wiped out several of my other datasets from my main pool before I stopped it. Again, no data loss, I have other sources. I confirmed in my console log, the command was "zfs destroy -Rr sr66-nas-v01/auto-20210622.0830-2w-clone"
Here are my datasets on this nas (after rebuilding, mid repopulate):
So, I recreated my datasets, repopulated my NAS, good to go. I still have this dataset I want to get rid of, but when I go in the GUI now and do the delete (it's there now) it tells me there are dependent datasets, and it's all my root datasets in the pool again.
Additionally, I purged the snapshots (I didn't need them on this NAS), however a single snapshot won't be deleted...the one I cloned to a new dataset.
Is there a safe way to delete the cloned snapshot dataset, and the 1 remaining snapshot?
I needed to pull a file I overwrote from a snapshot. I found the snapshot, in the gui I cloned it to a new dataset. I went in via ssh, copied what I needed out, then tried to delete the dataset, and the option was greyed out. I googled it, they said I had to promote the new dataset using "zfs promote" which I did, then the result said to do a zfs destroy of the dataset, which prompted for a -r and -R. I did that, and it wiped out several of my other datasets from my main pool before I stopped it. Again, no data loss, I have other sources. I confirmed in my console log, the command was "zfs destroy -Rr sr66-nas-v01/auto-20210622.0830-2w-clone"
Here are my datasets on this nas (after rebuilding, mid repopulate):
Code:
sr66-nas-2# zfs list | grep sr66 sr66-nas-v01 5.56T 11.9T 510M /mnt/sr66-nas-v01 sr66-nas-v01/.system 801M 11.9T 794M legacy sr66-nas-v01/.system/samba4 735K 11.9T 735K legacy sr66-nas-v01/.system/syslog-adb946163d914f088dc14617dbc0bec3 6.46M 11.9T 6.46M legacy sr66-nas-v01/Documents 66.1G 11.9T 66.1G /mnt/sr66-nas-v01/Documents sr66-nas-v01/Games 311G 11.9T 311G /mnt/sr66-nas-v01/Games sr66-nas-v01/HealthEducation 84.7G 11.9T 84.7G /mnt/sr66-nas-v01/HealthEducation sr66-nas-v01/Music 79.8G 11.9T 79.8G /mnt/sr66-nas-v01/Music sr66-nas-v01/Pictures 14.8G 11.9T 14.8G /mnt/sr66-nas-v01/Pictures sr66-nas-v01/TJGBackup 1.02G 11.9T 1.02G /mnt/sr66-nas-v01/TJGBackup sr66-nas-v01/Videos 5.02T 11.9T 5.02T /mnt/sr66-nas-v01/Videos sr66-nas-v01/auto-20210622.0830-2w-clone 510M 11.9T 509M /mnt/sr66-nas-v01/auto-20210622.0830-2w-clone sr66-nas-v01/bitcoinbackup 192K 11.9T 192K /mnt/sr66-nas-v01/bitcoinbackup sr66-nas-v01/litecoinbackup 192K 11.9T 192K /mnt/sr66-nas-v01/litecoinbackup
So, I recreated my datasets, repopulated my NAS, good to go. I still have this dataset I want to get rid of, but when I go in the GUI now and do the delete (it's there now) it tells me there are dependent datasets, and it's all my root datasets in the pool again.
Additionally, I purged the snapshots (I didn't need them on this NAS), however a single snapshot won't be deleted...the one I cloned to a new dataset.
Is there a safe way to delete the cloned snapshot dataset, and the 1 remaining snapshot?
