foo_fighter

I see similar behavior on CPU temps even after re-pasting. CPU fan is pwm2 I believe, but I've turned it off and just let BIOS control it.

I'm not sure of the syntax, but yeah the whole thing did work for me. I didn't try without the ghcr.io/

I used: ghcr.io/immich-app/postgres:16-vectorchord0.3.0-pgvectors0.3.0

You should try to get the s3_sleep plugin working.

You have a degenerate case, where in a 1 parity, 1 data disk system, the parity disk would contain identical bit-for-bit information as the data disk.

Assuming your windows dir is mounted as /mnt/windows/dir https://explainshell.com/explain?cmd=rsync+-avu+--delete+/mnt/windows/dir /mnt/user/backups/

FYI, updating to Hub 10 (V31) seems to have gone smoother than from V29 to V30 for me. ================================= Initializing nextcloud 31.0.0.18 ... Upgrading nextcloud from 30.0.6.2 ... .. [Fri Mar 14 13:16:43.945450 2025] [mpm_prefork:notice] [pid 1:tid 1] AH00163: Apache/2.4.62 (Debian) PHP/8.3.17 configured -- resuming normal operations [Fri Mar 14 13:16:43.945487 2025] [core:notice] [pid 1:tid 1] AH00094: Command line: 'apache2 -D FOREGROUND' ==================================

ZFS in the array is certainly viable. Having 1 or 2 ZFS disks allows for replication from cache(real-time backup of docker data without stopping containers) for example, plus the snapshotting, compression and scrubbing.

Set the version in the repository section: ghcr.io/blakeblackshear/frigate:0.15.0-rc2

Best case scenario is that they are exactly the same size or the new nvme is slightly bigger. If not, here is a post I made about a situation where the nvmes are slightly different sizes:

data-root is docker setting, either a vdisk image (docker.img) or a directory.

precleared devices don't have a filesystem. You have to format it to a chosen filesystem first and then you can mount it.

You could also use user.scripts and run it in the background with a log: Example: rsync -avuXizhP --size-only --delete --stats --log-file=/mnt/cache/rsync_logs/sync_Video_to_Tower_`date +%Y-%m-%d_%H.%M`.log --backup --backup-dir=/mnt/user/rsync_backup_dir/Video_`date +%Y-%m-%d_%H.%M` /mnt/user/Video [email protected]:/mnt/user

Unraid won't let you switch boot drives, only bios can do that, but there is a way to run a Win11 VM, from a bare metal install path as a VM. There is a YouTube video from SpaceInvaderOne on how to do it.

You can use a combination of CLI commands: zfs rename -p pool/old_dataset pool/new_dataset (must be in the same pool) or syncoid -r old_pool/old_dataset new_pool/new_dataset and zfs destroy pool/dataset to delete datasets to help speed things up.

So apparently it's not possible: Here's what I did: Run the user.script to snapshots/syncoid from cache to Array Stop Array Rename old cache pool to cache_old and mount with unassigned devices Create a new single disk cache pool with zfs as the file system Start Array Format new Cache to zfs Syncoid all datasets from cache_old to cache to preserve all snapshots: >syncoid -r cache_old cache rsync other files and non-datasets: root@NAS:/mnt/cache# rsync -avuXizPh --stats /mnt/disks/cache_old/ /mnt/cache sending incremental file list .d..tpog... ./ >f+++++++++ docker.img >f+++++++++ jwt-secret >f+++++++++ oplog.sqlite >f+++++++++ oplog.sqlite.lock >f+++++++++ rsync.log Double check all files made it safely Stop the array Set the number of slots to 2 for cache Assign cache_old to slot 2 of cache Note: it doesn't seem like you can select stripe or mirror like you can when you create a pool: vs but it eventually does show: and root@NAS:~# zpool status pool: cache state: ONLINE status: One or more devices is currently being resilvered. The pool will continue to function, possibly in a degraded state. action: Wait for the resilver to complete. scan: resilver in progress since Sat Jan 4 12:48:17 2025 593G / 593G scanned, 314G / 593G issued at 3.65G/s 317G resilvered, 52.98% done, 00:01:16 to go config: NAME STATE READ WRITE CKSUM cache ONLINE 0 0 0 mirror-0 ONLINE 0 0 0 nvme0n1p1 ONLINE 0 0 0 nvme1n1p1 ONLINE 0 0 0 (resilvering) and finally:

When you install new drives, sometime the BIOS re-orders them and messes up the boot order. See if you can access BIOS(Del or F2 when powering up), see all the drives and the USB disk and re-assign the USB disk as first in the boot order. If you cannot even access BIOS, that's a larger problem.

I have an existing single disk ZFS cache drive with a 4.1TB SP UD90 Nvme. I would like to add a second drive to create a mirror, ideally keeping the existing data. The new drives seem to be 4.0TB. Would this be possible? Here is some relevant data: Crucial P3 4TB: Nvme version: 1.4 Number of namespaces: 1 Namespace 1 size/capacity: 4,000,787,030,016 [4.00 TB] Namespace 1 formatted LBA size: 512 Namespace 1 IEEE EUI-64: 6479a7 844000005e SP UD90 4TB: Nvme version: 1.4 Number of namespaces: 1 Namespace 1 size/capacity: 4,096,805,658,624 [4.09 TB] Namespace 1 formatted LBA size: 512 Namespace 1 IEEE EUI-64: 6479a7 801a500c58

Is it possible to dynamically change the list sort order of the datasets? For example, by name, size, creation date, ascending or descending would be really nice too.

When you hover your mouse over them, it explains.... rollback hold release clone

I don't think I changed any other settings. I have this in my /boot/config/go file but I don't think it's needed: #Enable iGPU modprobe i915 chmod -R 777 /dev/dri Also in immich I have: And if you go in docker->right click immich-> >_console: root@60b2474b4ec9:/# ls /dev/dri card0 renderD128 root@60b2474b4ec9:/#

I think it's more a matter of colliding mount points as removing a disk from the array and then using Udev to mount it will mount it to the original location. You have to use a zfs command line to change the mountpoint. Creating a zfs disk in Udev also has some mount point problems once it's put into the array and then taken back out. Maybe these are pathological use cases.

spaceinvaderone's sanoid/syncoid user.script is pretty easy to use. For manual snapshotting zfs_master is also pretty convenient. A built in GUI would be nice to have.

Maybe chances are slim but if you had installed UnRaidConnect there is a backup of your config file online.

I ran zpool import disk7 and now it shows up in Udev as disk7: but I feel like there will be a conflict when or if I add a disk7 to the array esp a zfs disk since the pool names will collide. I think this is how I ran into a problem with missing a mount that I posted about in another thread.