T4ke

went back to btrfs for now, no more kernel panics. Solves the problem for me at least so I would state that its not a hardware issue in general.

Since I switched to a new platform, I have been experiencing kernel panics and call traces (see attached diagnostics). Old system: ASUS WS C422 PRO/SE (latest bios) Intel Xeon W2140B 64Gig DDR4 ECC Ram New system: ASUS Pro WS W790-ACE (latest bios) Intel Xeon w3-2425 64Gig DDR5 ECC Ram Hardware that I have kept from the old system: Nvidia Quadro P400 4 x WD SN850 Black, attached to an ASUS Hyper M.2 Quadro Gen5 Card Corsair Force MP600 ADATA SX6000LNP Intel X710 SFP+ NIC Broadcom / LSI 9400-16i HBA (IT mode) The old system ran flawlessly for 3 or 4 years with almost the same config, the new one has been in operation for about 4 weeks and constantly throws said kernel panics or call traces. ChatGPT says the kernel panic is in relation to ZFS modules and actually with the switch I moved from btrfs to ZFS for my cache / nvme drives. Is the platform still too "new", or what could be the reason? fuhlyserv01-diagnostics-20240111-1116.zip

Instead of downgrading your unRAID installation you could try my approach using a Linux VM with a Netatalk server. Works like a charm from my end. Currently backing up 4 Macs with this workaround.

Maybe there are some incompatibilities between Apples latest changes in their SMB stack and the different Samba versions of Unraid, that would be my first guess. In my statement I referred to postings in many different forums (Apple, Synology, iX Systems, etc), whose TM users all describe the same problem so I highly doubt that its a problem with Unraid. Most likely Apple is doing is own thing again, wouldn't be the first time.

It's pretty obvious that Apple made changes in the SMB implementation of macOS that seem to cause these issues. Its not Unraids nor Sambas fault. My temporary solution is to run a Netatalk/AFP server via a VM that gets the TimeMachine share passed through. With this, the incremental Time Machine backup works properly again. Also the already existing backups which were transferred before remain since only the network protocol is changed. I'm using Fedora since its using the latest version of netatalk which has some security fixes. This is only a temporary solution since AFP is insecure and was depreciated by Apple years ago but for home usage it's fine I guess.

how would I do that? Thank you for your answers so far, I already feared this.

Hi everyone, I have a general question about Unraid's Rebuild Process. First of all about the situation: I have 16 hard disks in my server. 8 are connected to the motherboard SATA controller, 8 more are connected to an Adaptec 71605 controller. All hard disks are XFS formatted and encrypted. Now last night I made a serious mistake: In the course of a firmware update I wanted to flash to the Adaptec controller the headers of all hard disks connected to the controller were overwritten up to the offset FFFF. Why this happened, I can not explain. I have not been able to find anything about whether others also had to make this experience in the course of a firmware update. In any case, these hard disks are no longer recognized by Unraid since the LUKS header was obviously destroyed. The other still working disks have their LUKS header at offset 8000 so it would be within the erased offeset range (don't know if Unraid genereally places the LUKS header at offset 8000). As I found out through my intensive research it's not possible to recover the LUKS header (or the data on that specific partition) without a binary header backup (which I don't have). Things I have already tried and and turned out not working: Search LUKS header using Hex Editor Clone LUKS header of another still working disk, hoping Unraid always uses the same master key within the array (which was not the case) Place a Loop Device on the LUKS device, starting at offset 8000. Anyway, I resigned myself to the fact that the data of the 8 hard disks is lost in a direct or indirect way. I do have a backup but it is no longer 100% up to date. My question now is as follows: Since I have two parity drives, would it be possible to recover at least two of the hard drives and leave the others "as it is" or does Unraid need the other drives to be intact to do this? If it's possible, how would I proceed here? Many thanks in advance and a wonderful christmas season, Robin