Search results

So an example. Assuming you using rpool for hosting virtual disks. rpool would have ashift=12 (4k block size), it would be 4k aligned, proxmox should handle this automatically in installer. Verifiable with tools like fdisk, sgdisk and zpool command. Then the zvol used to create the disk, would...

After some more thinking I actually think it doesnt matter providing the guest is writing in itself with 4k alignments, so 4k start (or multiple) for first partition and cluster size with a multiple of 4k. A bit like how it doesnt matter using a 512e disk on bare metal as long as 4k aligned.

Thank you, you was right. Seems I am learning all the time, and was something about UEFI I didnt previously know and now understand why Proxmox warns about a missing EFI disk. So the original install there was no Proxmox EFI disk, and Proxmox documentation says it makes a temporary one when it...

So this one has me bamboozled. I did a exotic setup in a VM mirroring one that is working fine on proxmox 7.3. Essentially the VM is running Debian bookworm, and has a EFI bootloader on a EFI partition. When trying to boot from the VM configured to use UEFI, it will just say a vague it cannot...

These are the type of things I want exposed in the proxmox gui settings, thank you for revealing this, also like the scan zeroes setting that is forcefully enabled when enabling discard in proxmox, that should be optional. https://bugzilla.proxmox.com/show_bug.cgi?id=4503.

Thank you for this, it seems they have changed the zfs structure in Proxmox 8 They now added 'rpool/var-lib-vz' as a separate data set, which had caused 2 datasets I mount under images to block the boot process. Interesting that if a non critical mount fails the entire boot process is killed...

By default some form of THP is used, I know as I felt the downsides a while back, was getting OOM's forcefully shutting down VM's (host had 32 gigs of RAM, with several gigs free when this occurred). Setting THP to never on the kernel went a long way to stabilising the issue, also is this post...

I would keep that just in case you need to boot via legacy boot. Partition table needs to be GPT format for UEFI to work. Older Hetzner servers dont support UEFI booting out of the box, so be aware of that, the good news is if you on such a server you can request to switch to UEFI mode on a...

Fighting this right now, I have done the install via qemu using my normal method, but the boot fails at post (looking via KVM) as if there is no OS installed, however it will boot fine inside a qemu instance inside rescue os. -- Fixed it via setting storage oprom to UEFI only in bios. So...

I will do the request if it hasnt been done already, I think its really wasteful of nand writes and adds an extra delay to use /var/tmp as a staging area for ISO's. in the mean time I might mount /var/tmp as a ram disk.

Should i remove 6.1 and 5.19 kernels before installing this update then, as you state its 5.15 or 6.2 only?

https://www.kernel.org/doc/html/latest/admin-guide/kernel-parameters.html?highlight=amd_iommu As you said the on parameter isnt a valid option for it, so thanks for pointing that out. So my iommu works in its default state, 'pgtbl_v1' the default configuration. Maybe I got the syntax from...

There is a second way to enable IOMMU, isnt well known, but ASRock did a custom bios for me so I could have it working on a G chip. Until I next go in my bios I cant help on the setting (they eventually made it public, but not all of their boards). I also have these 2 in my cmdline, but not...

All my testing has been done on either single drive ZFS or mirrors. Increasing volblocksize decreases metadata and increases compression. But also increases waste if using small file sizes. Only few days ago I made a change on a windows guest. With 4k NTFS on 64k volblocksize it couldnt...

Considering using this over stock 5.15, I checked the compile config, 5.15 is set to preempt voluntary which I think is probably the best preemption mode, but 6.1 has full and both voluntary configured from what I can tell so unsure what its default run mode would be. I assume adding this to...

Well Nvidia seem to have released something to allow it to work over RDP for geforce cards, but its not clear if its the host or the machine running the RDP client that needs the Nvidia GPU, if its the latter it could be perfect.

Yeah I avoided doing that for now due to the dependency bloat, but since it seems to be the only GL option in Proxmox I will probably do the same when I have further time to test. Your feedback is good as it shows its probably worth it with the improvements you have, one issue though is there...

Its been a while, but thats because time is needed really for this stuff, I havent yet seen signs of new corruption, and have assigned io_uring now to multiple drives. So either its solved or for whatever reason it went away, will never know I guess, but will report back if its a problem again.

Well now I have a easy way of opting out its not a big deal. :)

I discovered that virtio gpu used by proxmox is actually virtio-vga (the two are different, so they perhaps should fix the naming to avoid confusion) I then changed it to virtio-vga-gl which is supported in qemu using custom args flags, but got a no opengl support error, and since I am busy had...