Search results

Meh. Sending ZFS snapshots to a safe place is nearly trivially easy if you care about restoring individual nodes.

I ran in to this a while back. Here was what I did to get around it: https://forum.proxmox.com/threads/proxmox-7-1-2-is-unable-to-install-on-amd-processor.116549/#post-514145 Pretty much muck around in the console after proxmox fails to start X and then add the bits and restart the process

Had the same sort of error today. 5.15.107-1-pve -J

Thanks for this. Ran into this yesterday and today, and nice to find the fix here.

That's what I was afraid of. Guess I'll have to get creative to find the info. Appreciate the responses. -Josh

Thanks for the response. Optimally, I'd like to get a list of what files were recorded as having changed, so that I can potentially move them to a not-backed-up second drive I use for cache and temporary folders. Stuff like log files in odd places, or other transitory things which don't need...

Is there a way to compare what changed between PBS incremental backups? I'd like to see what's changing in a Proxmox VE snapshot backup. Thanks!

I did the fix here: https://bugzilla.proxmox.com/show_bug.cgi?id=4230#c4 After it failed to boot, I just did the below instructions, no change to the boot line after boot and X11 fail to start - add Section "Device" Identifier "Card0" Driver "fbdev" EndSection (without BusID)...

Just tried this & ran in to the same thing. Are you using a GPU or the onboard graphics? I suspect the onboard graphics are not yet supported by the kernel/installer. I'll track down a graphics card and see if that makes a difference.

@chencho - I don't see an IP Address set for vmbr0 Here's my current config that now works. iface eno1 inet manual auto vmbr0 iface vmbr0 inet static address 144.###.###.### netmask 255.255.255.0 gateway 144.###.###.254 bridge_ports eno1 bridge_stp...

My one fresh install - from a self-uploaded ISO worked fine. Upgraded didn't - just like yours. OVH has protections to keep spoofing, so when your vmbr? interface had a different HW id, that probably caused OVH to not pass traffic. In my interfaces file, I added "hwaddress ether <mac...

@plokko Any chance you're at OVH? I had to do the same thing on an upgraded server there, and this may have to do with their Failover IP weird networking requirements.

Yep, it's a good one. For some reason, my vmbr0 did not have the same HW address as eno1. Changing them to match (and hard setting it in the interfaces(.new) file worked.) Looks like https://forum.proxmox.com/threads/pve-6-4-to-7-no-network-only-for-host.98198/post-425362 found the same thing...

And kernel 5.11.22-4 and 5.11.22-5 made no difference. So after a time, on a wholly working system using the original OVH Proxmox 5 or 6 template, the host OS loses its connection to the OVH network beyond the local /24. Virtual Machines continue to work. (I've seen this on two systems.) No...

The network interface on my server just hung again, changing to 5.11.22-4 to see if it behaves.

I am in my case. It works, and then the host (only) stops talking on the net. I'll know soon if it's a -5 idle thing.

What kernel version? I'm troubleshooting a similar issue where the Proxmox host loses connectivity to the public Interface, but VMs are fine. I have a suspicion that it's a bug in 5.11.22-5-pve because the server I have running 5.11.22-4-pve is okay. Maybe try installing...

What's the underlying OS, and what processor settings do you have for the VM? I've found that it's best to power down and restart a VM when going from AMD <-> Intel.

Check out this section of the wiki -> Limit ZFS Memory Usage https://pve.proxmox.com/wiki/ZFS_on_Linux

I'd guess memory contention. Turn off ksmd or at least do a: echo 2 > /sys/kernel/mm/ksm/run