I am currently running a 12TB TrueNAS system, primarily as a media server but also to back up other systems in the house. I found a good deal on a 14TB external USB drive, so I decided a proper backup for the NAS was a good idea. I set it up according to recommendations in this thread: Reddit - Dive into anything (specifically the recommendation by konzty to use a single-disc pool on the backup drive, and then use Replication to regularly snapshot the changes and copy to the backup. Everything seems to be working, I get notifications that the replication tasks have succeeded, etc. Howeverā¦
My dashboard for the backup pool says that there are over 12 TB free, and Iāve effectively used 0% of available space. That definitely doesnāt seem right! If I go look at pools, for instance, it shows similar info. Any ideas? Is TrueNAS somehow not reporting the state of the backup drive correctly? Seems more likely that Iāve set up something incorrectly and it never copied the initial image, but it seems relatively simple and Iām not sure what I could have done wrong (I am in no way an expert on this stuff however!).
Iām not even sure how to check the content on the backup drive outsid3e of TrueNAS, since I donāt think any of my other machines can understand a ZFS disk.
You probably replicated the top-level root dataset, without invoking ārecursiveā or āfull filesystem backupā in the Replication Taskās options.
Without seeing any more information, we can only speculateā¦
So the zpool list shows all space free on the backup drive. But I think the suggestion above that neither Recursive nor Full Filesystem Backup were checked is correct, I just checked my task and indeed they are both unchecked.
Full Filesystem Backup seems pretty obvious, but Iām not clear about the Recursive. Is that something that I should check for a backup replication task?
āRecursiveā will include all child datasets that live underneath the selected source dataset. (Since you chose the root dataset as the āsourceā, you basically selected an empty āplaceholderā dataset. Your data lives on children dataset.)
āFull Filesystem Replicationā implies recursive (above) + all properties, volumes, and clones.
OK, sorry for the newb questionsā¦I checked the full file system replication and re-ran it, and quickly go the following error notification (Note: ādeadpoolā is the main data pool Iām trying to back up):
Replication ādeadpool - NAS Backup Poolā failed: skipping snapshot deadpool@auto-2024-06-04_00-00 because it was created after the destination snapshot (auto-2024-06-03_00-00) skipping snapshot deadpool@auto-2024-06-05_00-00 because it was created after the destination snapshot (auto-2024-06-03_00-00) cannot send deadpool@auto-2024-06-03_00-00 recursively: snapshot deadpool/iocage@auto-2024-06-03_00-00 does not exist warning: cannot send ādeadpool@auto-2024-06-03_00-00ā: backup failed cannot receive: failed to read from streamā¦
Update: I recreated the whole thing over and itās definitely creating snapshots every day like I scheduled, and the disk is now showing like 70% full, which is what I would expect from backing up the whole system. It is telling me that the backup pool is ādegraded.ā Itās a brand new drive so that seems odd. Not really sure what that means.
I still need to try to poke around on the backup drive and see if the files seem correct and uncorrupted.
Just to give a (hopefully) final update on this. I added a USB 3.0 card, removed the backup pool and started it again with the USB 3.0 interface, and so far there have been no issues, no mysterious replication failures or degraded pools, and of course backups are much, much faster. Fingers crossedā¦
