trurl

How are you testing? The speed you mention seems typical for parity writes so if your test is transferring files to the parity array your NIC isn't really the reason. On mobile now so can't look at Diagnostics.

A better way to say it is you should not mix user shares with disks that are included in user shares. If your ZFS disks are not part of the array or cache they are not included in user shares.

Yes, parity plus all remaining disks allow the data for a missing disk to be calculated. Parity is a common concept in computers and communication, and its meaning is basically the same wherever it is used. Parity is just an extra bit that allows a missing bit to be calculated from all the other bits. Here is the wiki on Unraid parity: https://wiki.unraid.net/UnRAID_6/Overview#Parity-Protected_Array

Parity isn't a backup, so whether or not you have valid parity, you must have another copy of anything important and irreplaceable, preferably on another system.

But mounting an array disk outside the array will invalidate parity whether or not the array is stopped. The removed disk is modified outside the array and no longer matches parity, so parity is invalid. And when you put it back and start the array it will look like everything is OK since Unraid doesn't really know what you might have done with the array stopped. So then when you have a problem that requires valid parity to recover, it won't work because that disk is out-of-sync. Parity by itself can't recover anything. All other disks are also required by the parity calculation.

no I backup critical files to Unassigned Devices for storage offsite. Scripts can be run automatically when Unassigned Devices are mounted. There are examples on the first page of the Unassigned Devices plugin page. There are lots of different dockers for backing up and/or synching to different things. Really a simple rsync script will do for a lot of purposes. And there is a User Scripts plugin to help with running scripts.

Until you get more experience, please ask for advice before making any changes to disk assignments. Removing disks and mounting them outside the array, for example, will invalidate parity. Sounds like parity wasn't valid yet anyway, but just thought I would mention it since it seems like something you might do without asking for advice. When syslog starts, you don't have anything assigned as parity or disk1. I guess we will just have to assume disk1 is OK, or at least as good as it we can make it, since you may not have valid parity. I assume this is the disk you were going to use as parity: Serial Number: ZL23XXAR Go to Tools - New Config. Assign all disks including parity exactly as before. You must be very careful and not accidentally assign a data disk to the parity slot, or the data will get overwritten by parity. Then start the array and let parity build. I recommend waiting until you have valid parity before using the array just in case you have any other problems. 16TB may take more than a day.

Unassign disk4 Start array with disk4 unassigned Stop array Reassign disk4 Start array to begin disk4 rebuild

CA Backup plugin will backup appdata, libvirt.img, and flash. docker.img backup isn't needed since the actual docker executables are easily downloaded again. So that mostly takes care of appdata and system shares. Don't know what else you might be keeping on cache. I have subsets of my music and photos on mine for faster access by other devices on the network, for example, but these are just copies of things already on the array. Search for backup on the Apps page for other ideas.

Do you actually want disk4? Sounds like you might have been working towards removing it.

That seems possible. There are no signs of any problems otherwise.

Looks like disk4 is the one that is disabled and needs to be rebuilt. But neither disk is showing in SMART. Do you have any reason to think they are both dead? Check all connections, SATA and power, both ends, including splitters. Then post new diagnostics.

Since you only have single parity, you can only recover 1 missing disk. You have 2 missing disks. Doesn't matter whether or not there is data on any of them. How did you get to a place where you have 2 missing disks? Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

Did you have any other problems before you noticed the files were missing that you tried to fix yourself?

Sorry, of course you can't work with Krusader with docker disabled Just go to the command line and rm -R /mnt/disk1/system Then go to Settings - Docker and enable dockers. Then the Previous Apps feature will reinstall your dockers exactly as they were.

Not sure how you ever got libvirt created then. Do you know how to work directly with the disk files and folders at the command line or with mc (Midnight Commander) or Krusader? Basically you need to get rid of the system folder on disk1.

Go to Settings - Docker, disable, then delete docker image from that same page. Do the same for Settings - VM Manager. Don't enable either for now. Then post the results of

Just assign the new disk to the parity slot and start the array to build parity

Do you have any VMs?

Replacing disks won't help. Get an UPS. Leave as Prefer. Disable Dockers in Settings then run Mover

Diagnostics already includes syslog. The only time we might want a separate syslog is if you have Syslog Server enabled to get us syslog from before a crash. appdata is all on cache now, but system still has files on disk1. Do you mean disabled the docker container you had running, or do you mean you disabled the docker service in Settings - Docker? The service must be disabled to move system. What do you get from the command line with ls -lah /mnt/disk1/system and ls -lah /mnt/cache/system

Nothing obvious there and nothing in syslog really after parity copy begins. Can you tell us more about the hardware?

CRC errors are not disk errors, they are connection problems. According to your syslog you may also be having connection problems with cache. Check all connections, SATA and power, both ends, including splitters. CRC errors won't reset, but you can acknowledge them on the Dashboard by clicking on the warning, and it will not warn again unless they increase. Also, your appdata has files on the array.

No Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

Since these shares are cache-prefer (by default) new files will always go to cache if there is room. But if they have to go to the array. Since these are cache-prefer shares mover will try to move them to cache, but then go to Main - Array Operations and Move Now. Also, mover will not move duplicates, so post new diagnostics when it completes so we can see if there is anything else that needs to be cleaned up