Search results

# for d in /sys/kernel/iommu_groups/*/devices/*; do n=${d#*/iommu_groups/*}; n=${n%%/*}; printf 'IOMMU group %s ' "$n"; lspci -nns "${d##*/}"; done IOMMU group 0 00:00.0 Host bridge [0600]: Intel Corporation Xeon E5/Core i7 DMI2 [8086:3c00] (rev 07) IOMMU group 10 00:04.6 System peripheral...

I don't know. How can I find out? Booting from a USB stick could be a possibility, though I had a quick google and it does seem a bit fiddly. Maybe just buy a 5 port PCIe adaptor, amazon have a couple different models from around €50 - would something like this work with Proxmox?

Yes, I think they are all there: # find /sys/kernel/iommu_groups/ -type l | grep "02:00\|00:1f" /sys/kernel/iommu_groups/25/devices/0000:00:1f.2 /sys/kernel/iommu_groups/25/devices/0000:00:1f.0 /sys/kernel/iommu_groups/25/devices/0000:00:1f.5 /sys/kernel/iommu_groups/25/devices/0000:00:1f.3...

I have a HP Z420 with 8 SATA drive connectors. According to the documentation it is 2 x 2 AHCI and 1 x 4 SCU, see diagram below. Currently I have a system drive on (15), 4 of 5 raidz drives on (17) and the last raidz drive on (14). I'd like to passthru the raidz drives to a VM, so I'm trying...

Right, I did mean TrueNAS Scale. I think it's relatively stable, the kernel is Linux mostly it's the UI that's beta, and the community seems to suggest it's stable enough for home use. Yes that would sound like a serious showstopper if you had to start all over every time there's an update...

Were you able to run TrueNAS in a LXC container, or did I misunderstand? I'm kinda considering going the opposite way, and replace Proxmox on my home server with TrueNAS scale. Effectively I only use it as a NAS and Docker server anyway, so might as well de-complicate matters. I'd like to...

Right, but what happened was the server was rebooted and a USB backup drive that normally uses /dev/sdg suddenly decided to use /dev/sdc instead, which caused one of the zraid drives to disappear from the pool, and so on and so forth. How could that happen and how do I avoid it happening...

Actually, isn't the pool already using by-id ? NAME STATE READ WRITE CKSUM poolz ONLINE 0 0 0 raidz1-0 ONLINE 0 0 0 wwn-0x50014ee20fb7a683 ONLINE 0 0...

Still can't export the pool, keep getting the same cannot export 'poolz: pool is busy" So, under Datacenter -> Storage I have 3 entries: It turns out all of those hook into /poolz, so I disabled all 3 and rebooted. When the server came back the pool was gone. # zpool list no pools...

And how do you do that, just a standard 'ls -alR' or 'find' showing nothing, or is there some zpool export-check command?

How do you do that? Just like any other folder or is there a zpool command to remove it?

Ok I will try that. Assuming I manage to export the pool, I understand from the manpage that zpool import will want to use a new pool name. I'd prefer using the same old - is that possible? How? Once the pool is exported, can the old pool be deleted or something?

it's added as local-backup storage yes. (VZDump backup file)

After having had a boot problem because I'd used device names for my zfs pool rather than ID's, I realize I have to change it. Searching for hours, all I find are a couple of references that basically says to "just" ... zpool export storage zpool import storage -d /dev/disk/by-id When I try...

Hm, this is interesting. Seems the second drive in the raidz array had been knocked out by a USB attached backup drive that wanted to use the same drive id. I removed the USB drive and restarted the server and then the missing drive was there and the pool could get started. After that the...

Today I shut down my PVE, apparently without problems, but when I switched it on again half an hour later I got a lot of zfs errors, and zpool status now shows longer and longer time "to go": # zpool status pool: poolz state: DEGRADED status: One or more devices is currently being...

Thanks, yes that makes sense. So passing through a disk drive this way, other than limiting access to things like SMART, is there a noticeable performance loss with this? The plan is to pass through 5 drives for a zfs pool for a Truenas VM.

If I have a 2 (or more) port PCI controller on the server, is it possible to pass it through to a VM so the VM can use one disk on the controller, while the host is using another disk on the controller? I could of course just try it, but since the host disk is the boot disk it might cause the...

I know ZFS isn't a proxmox specific feature, but the version used in proxmox is different from the one documented in Oracle. Anyway, this irrelevant point apart, when I try zfs change-key, I get this error: # zfs change-key -i poolz/secdata Key change error: Parent must be encrypted. If I...

Similar problem here. After upgrading to Proxmox 7, docker in a CT no longer starts. I find with docker problems, running dockerd without options gives the most usable debugging output: root@docker ~# /usr/sbin/dockerd INFO[2021-10-24T11:58:17.515590860Z] libcontainerd: started new...