nimble_bitz

Excellent thanks for the reply.

Is it at purchase or when activated as a key? If I buy for a buddy and he doesn't build for 6 months did he burn the time? The black Friday/Cyber Monday has me itching to get him into the hobby.

I would recommend shutting down. Double check cabling, and then restart the array and attempt parity build again. If it continues, post your diagnostics. Quick question how is your new parity attached to your motherboard? Direct to a port on the board or through a riser card or backplane?

Are any docker's using the GPU? I can't remember if the Epyc chips had integrated graphics. So if docker is using it you have to unmap that as it can't live in both worlds local to the kernel and as passthrough to a VM. If not verify passthrough in the tools> system devices and check the boxes next to the card and card audio. Reboot I think is required afterwards. Word of warning, you may not get a display until you assign to the VM and turn the VM on. IOMMU passthrough without a secondary GPU can be "weird" for lack of better words.

Remember if you swapped your USB in under 1 year you'll have to contact support for help with the key activation against the new GUID.

The squashfs is the liveusb feature (the booted USB). It's having issues writing. Dumb question when you swapped drives, did your full USB get copied over or just the config folder? Slight chance a bad interaction and more likely the flash drive.

Thank you much for the advice.

Worth the effort to swap again? I have another spare I could throw in. Was having a hell of a time getting the array up with the swapped drive due to the L2arc SSD issue. Picked less of 2 evils at the time. Protect my data or remove the SSD that croaked.

Was done through cli, drive was a dead number and ran a zpool replace zpool NUMBER sdb (as the /dev/ didn't work) Original was encrypted, not sure why this one didn't. tower-backup-diagnostics-20241127-1140.zip

So when I replaced a failed drive it resilvered however it doesn't look right. /mapper/ is missing, how did this happen and how do I correct? Relatively new to zfs even though my day job is a NAS administrator, more accustomed to WAFL.

Got it had to remove the /dev/ then it worked. Any idea how to fix the issue with the /dev/mapping/sdb1 actually being /dev/sdb ? I can make a new topic if needed, but your help is much appreciated!

Okay I got the order to change so /dev/sdg1 doesn't exist in the array as it's an assignment that can move. However I can't remove /dev/sdg1 pool: zpool state: ONLINE status: One or more devices could not be used because the label is missing or invalid. Sufficient replicas exist for the pool to continue functioning in a degraded state. action: Replace the device using 'zpool replace'. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J scan: scrub canceled on Wed Nov 27 07:04:16 2024 config: NAME STATE READ WRITE CKSUM zpool ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 sdc1 ONLINE 0 0 0 sdd1 ONLINE 0 0 0 sdb ONLINE 0 0 0 sdf1 ONLINE 0 0 0 sde1 ONLINE 0 0 0 cache sdg1 UNAVAIL 0 0 0 errors: No known data errors root@Tower-backup:~# zpool remove zpool /dev/sdg1 cannot remove /dev/sdg1: no such device in pool

Do you have an example syntax to use? Where I run into concern is this: zpool ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 sdd1 ONLINE 0 0 0 sde1 ONLINE 0 0 0 sdc ONLINE 0 0 0 sdf1 ONLINE 0 0 0 sdg1 ONLINE 0 0 0 <--- (actual array drive) cache sdg1 FAULTED 0 0 0 corrupted data (was bad ssd but it assigned to the same device number)

Created a subpool and I can't get rid of it, as the dev identified as another device when the ssd failed. Need to remove this to properly fix a device rebuild of sdc which should be sdc1 but it was brought in weird. Would love some help as I'm new to zfs. root@Tower-backup:~# zpool status -v pool: cache state: ONLINE scan: scrub canceled on Tue Nov 26 16:25:47 2024 config: NAME STATE READ WRITE CKSUM cache ONLINE 0 0 0 nvme0n1p1 ONLINE 0 0 0 errors: No known data errors pool: zpool state: ONLINE status: One or more devices could not be used because the label is missing or invalid. Sufficient replicas exist for the pool to continue functioning in a degraded state. action: Replace the device using 'zpool replace'. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-4J scan: scrub canceled on Wed Nov 27 07:04:16 2024 config: NAME STATE READ WRITE CKSUM zpool ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 sdd1 ONLINE 0 0 0 sde1 ONLINE 0 0 0 sdc ONLINE 0 0 0 sdf1 ONLINE 0 0 0 sdg1 ONLINE 0 0 0 cache sdg1 FAULTED 0 0 0 corrupted data

Anyone else having issues with ARC and Tone mapping hardware transcoding with 7.0 beta2? I've added in --device=/dev/dri:/dev/dri also did the Device passthrough. All other types of transcoding work and 4k works with hardware transcoding with tone mapping turned off.

Seeing a similar issue using ARC 380 and the new 7.0 beta. Tone mapping breaks the hw transcoding.

Was there plans for a release updated for 6.8 stable?

Quick question, were there changes in this build that affected the br0 settings with VMs? I can't isolate out VPN vs no VPN traffic without using the local client now. Used to just add the IPs to an alias list that routed it on the individual br0 IPs. Pfsense gateway routing used to work for unraid now doesn't function unless I remove the unraid server IP from the VPN. For clarification I haven't tested against the Limetech build as I just figured this out yesterday. Might be worth some testing.

This worked beautifully on my msi x370 sli plus that I updated for ryzen 3000. Was pulling my hair out trying different settings to get iommu working until I found this thread. Mucho kudos, now if we could get Limetech to implement these fixes....

Been a user since 2014 I believe. Got started with RCing 6.0. Original plan was to createna data location that could house my family's vast picture collection off my main rig. It has morphed into so much more. Virtualization, docker, etc. Own 2 licenses and recommend to all my coworkers.

If you have satellite DECA is an option as well. It’s Directv’s version of MOCA as satellite runs in a different frequency range.

Here’s my AIO rig. 34TB Array, 2 480GB cache and a 512 NVME running my vm Win 10 in unassigned devices doing IOMMU passthrough. I know, I have to reroute the USB front header cable or move to the last 1x PCI-E. SpaceInvaderOne inspired me with the videos. This is in a cooler master mastercase maker 5. Still have space for 2 more under the shroud by PSU.

Figured it out 900K entries that needed to be deleted from Recycle bin manually due to a rouge creation of Squid maleware bait files must’ve crashed it. Once I cleared them the plugin started working again. Must’ve been too much scrub for the plugin.

Yay I’m not losing my mind. Mine is doing the same thing on 6.5.2