Search results

How do you mean forgotten? Did you set 2-2 and it returned to 1-2, or did you set 1-2 and the device now has 2-2? Could you post your VM config (both working and broken, if there is a difference) via 'qm config <vmid>'?

With kvm64 you do not get nested virtualization at all... Could you explain your setup a bit more, what are your L0, L1 and L2 guests and post all relevant VM configs? ('qm config <vmid>')

Der Kernel ist das low-level Interface zur hardware, das hilft bei storage replication so gut wie gar nichts. Ich muss zugeben, mich mit OMV nicht allzu gut auszukennen, aber da gibt es doch sicher bereits eine integrierte Möglichkeit für Replikation? Das wäre dann ja auch per GUI managebar. PVE...

As I already answered in the post you linked, this is something we'll look at in the future - so no, right now that's not possible. If you mean the config in the backup itself - also no, as backups are by definition immutable.

Nein, ist es nicht. Die 'zfs' Befehle werden bei "ZFS over iSCSI" per SSH am target ausgeführt, dafür setzen wir ja auch voraus, dass das 'zfs' binary und ein iSCSI target dort verfügbar sind. pve-zsync wäre u.U. noch denkbar, müsste allerdings dann auch dort installiert sein - vielleicht in...

pve-zsync/replication basiert auf lokalen ZFS pools - die sind bei iSCSI aber nicht vorhanden, das ZFS liegt ja am target. Dementsprechend müsste eine replikation auch dort eingerichtet werden - was ein wenig außerhalb des scopes für PVE liegt, auf dem target kann ja jedes beliebige system...

Thin or thick provisioning depends on the underlying storage, not any settings on the VM. How are you making the backup? Are you using compression? It would also help to post your VM and storage config ('/etc/pve/qemu-server/<vmid>.conf' and '/etc/pve/storage.cfg' respectively).

Only if you have a heterogenous cluster, so mixed Intel and AMD. Otherwise going for the lowest common denominator of whichever manufacturer you're using is usually better for performance while still having full compatibility. Mixed Intel/AMD isn't really supported for live-migration anyway...

Bei USB passthrough wahrscheinlich eher nicht. Unter Linux an sich geht das normalerweise schon (siehe z.B.: https://askubuntu.com/questions/56270/how-can-i-spin-down-external-hard-drive), aber beim passthrough wird das schwieriger, weil Windows da zwar Daten durchreichen kann, aber u.U. keine...

That already happens, automatically. What you're referring to is called "overprovisioning", i.e. you assign more cores to your VMs than you physically have. Then, when a VM isn't using some of it's cores, the performance will be available to others. In your example, you could assign your 4 VMs...

The set is limited to a whitelist because there are some CPU flags that can have security implications - AVX256/512 for example can stress the CPU very intensely, leading often to downclocking behaviour, and thus can cause performance issues on other VMs, even when scheduled to different cores...

Yes, accessing a file/directory on a bad NFS mount causes the calling process to lock up (since the syscall hangs). This manifests itself as a hanging pvestatd on that node, meaning it can no longer provide info for any of the available storages.

This looks more like data/storage corruption than a VM issue... The drive should be available on the host at /dev/mapper/pve-vm--100--disk--0, maybe see if there's any data in there? Did you maybe upgrade/change any storage parts you are using during the upgrade? Did it work before, did the...

Hm, this should work fine then. Just to confirm, could you try removing the 'sata0' disk from boot order, clicking "OK" and then adding it back? That should change the config to use the new format as well. "Cannot find the harddrive" as in: it doesn't show up at all - or you don't see any data...

Can you post your VM config? (qm config <vmid>) I suspect what happened is that you require multiple disks for booting, which in newer versions of PVE you have to *all* set as bootable. To do so, go to "Options" and open the "Boot Order" menu, in which you have to select the checkmark for all...

Are you using the same storage setup on both nodes? Otherwise you need to ensure that each storage is configured to only be available on the correct node in "Datacenter -> Storage". Note that when joining a node to a cluster it's own storage config is overwritten, so if, say, the second node is...

I can reproduce the behaviour consistently here. Still investigating the root-cause and a real fix, for now I've sent a patch to our mailing list to simply disable io_uring for specifically this scenario: https://lists.proxmox.com/pipermail/pve-devel/2021-June/049120.html

Can you post your storage config (/etc/pve/storage.cfg)? Most importantly from that, do you use krbd for your "CephRBD" storage, and does it make a difference if you change it?

We have extensive documentation on how to remove a node available: https://pve.proxmox.com/pve-docs/chapter-pvecm.html#_remove_a_cluster_node The reason is that removing a node is a somewhat dangerous operation that might incur serious data loss or downtime if executed incorrectly (e.g. the...

Yes, that card should be supported. Note that if you only have a single VM you can use the regular NVIDIA driver with PCIe passthrough, but for using the card to accelerate multiple VMs, you need an NVIDIA GRID license as far as I know. See NVIDIA's info about GRID for more about that, it is...