Smarter mover

[DarkKnight]

Why doesn't the mover script error out on it's own and leave a message in the GUI when there is no space left on a given device? It's currently stuck, and there doesn't seem to be a way out of it short of restarting my server. Some place in the GUI where we could watch the status of the mover would also be helpful. I'm not afraid of the CLI at all, but this is an important background function of Unraid, so viewing it's status (beyond 'mover running') should probably be accessible from somewhere in the GUI.

 

Related problem, replacing BTRFS cache pool drive I did a while ago also changed the raid mode to 1. I changed it back to 10 and did a balance. The file system indicates that there are ~2.7TB used and ~9TB free, but anything that tries to write to the cache pool gets a 'device full' error. 2.7TB is suspiciously equivalent to the total space of a given single 3TB drive from that pool of 8x 3TB drives (e.g. Raid1). In the GUI for the BTRFS pool it reads: Data, RAID10: total=2.77TiB, used=2.74TiB, but df -h reads: /dev/sde1  12T  2.8T  8.2T  26% /mnt/r10_cache. So, something is really wrong here. 

 

I don't really have the inclination to deep dive into why this happened. I'll just say it's growing pains for a new feature. My intention is to 1) never buy drives off of facebook again, and 2) recreate the cache pool as soon as I can move the data off. It would be cool if there was just a 'dump cache to array' tool instead of dicking with the cache settings and having it completely (and silently) malfunction for 2 days because the data moved to the cache instead the array.

 

Edit: I apologize for any snark. It's late, I'm tired and have a cold. 

Edited November 16, 2022 by DarkKnight

[ChatNoir]

2 minutes ago, DarkKnight said:

Related problem, replacing BTRFS cache pool drive I did a while ago also changed the raid mode to 1. I changed it back to 10 and did a balance. The file system indicates that there are ~2.7TB used and ~9TB free, but anything that tries to write to the cache pool gets a 'device full' error. 2.7TB is suspiciously equivalent to the total space of a given single 3TB drive from that pool of 8x 3TB drives (e.g. Raid1). In the GUI for the BTRFS pool it reads: Data, RAID10: total=2.77TiB, used=2.74TiB, but df -h reads: /dev/sde1  12T  2.8T  8.2T  26% /mnt/r10_cache. So, something is really wrong here. 

Please attach your diagnostics to your next post.