Hi, @sebalinux .
I know it would be a workaround, not a solution, but to locate the reason closer, you may try as well installing PVE 8 and then upgrading to PVE 9 and checking if it boots correctly.
I'm using the third-party proxmox-csi plugin for Kubernetes to create VM disks. As of the new version, I can't create new disks anymore. I get a 500 internal server error from the Proxmox API. I have been looking into the logs for the csi plugin...
Alright, well I ended up creating a whole project around this request, starting from the nice work that Jorge did. It just needed a bit more work and features for my use case (which is deploying OpenShfit/OpenStack for development and using the...
A quick note on when you want multiple encrypted ZFS pools in addition to having a ZFS system pool:
Turns out when you have more than one and try to create a second systemd unit for them you will cause a systemd dependency cycle, and things will...
First you need to check the performance of your storage, yo can test it with fio command
https://pve.proxmox.com/wiki/Benchmarking_Storage
Also you can try
pveperf
For example
pveperf /var/lib/vz
Is not a good idea making backups in different datastores as you miss the dirty bit, so every time you do a backup weekly or monthly is a new copy not incremental.
I suggest once again that people evaluate the Pulse monitoring app, which uses the Proxmox api to monitor not only the temps, but the CPU/RAM/Disk/Backups/hosts/docker apps etc. on a single pane of glass, right across your estate. They have made...
It does not appear to be retrieving incorrect values.
The raw value appears to be 79.
The normalized value is 96, and the 4% loss is correct.
We will verify the normalized rather than the raw value.
The reason drivestatus.sh retrieves raw...
Thank you uzumo... I just plugged in what you said into ChatGPT and I learned that the
> Samsung SMART RAW_VALUE=79 is not 79% wear
and that I should be going by the normalized Value...
I'll look into changing that...
Thank you for sharing...
Good update!
Some observations:
- As noted above the upgrade instructions are missing an apt source file example for the no-subscription repository.
- The notes mention kernel 6.17, however I'm only getting the normal 6.12.
- The FAQ mentions...
As I hate when people never really write what the solution was and just abandon a thread:
I solved my issue as followed (credit goes to https://blog.kye.dev/proxmox-zfs-mounts ofcourse)
Create a user and group on Proxmox host:
groupadd -g...
I suggest once again that people evaluate the Pulse monitoring app, which uses the Proxmox api to monitor not only the temps, but the CPU/RAM/Disk/Backups/hosts/docker apps etc. on a single pane of glass, right across your estate. They have made...
I checked and there is no Samsung Magician tool available for Samsung Evo Drives...
I even tried to download the only Linux tool I saw under Samsung Magician software (even though it said it's only available for DC drives) and predictably it...
Same exact problem here with latest rev. 9.1.1. By the way I got exact same error after luanching qm migrate from an old node version 8.1 to migrate vm to a new cluster (this one with 9.1.1)... I was initially thinking it was due to EFI disk and...
I have given up on getting lxc.idmap to work.
I found a working method via this post: https://forum.proxmox.com/threads/zfs-permission-and-ownership-problems.166084/post-771442
Correct, Homepage does not verify the certificate. And the cited error looks like the authentication fails. This may well be because the permissions are insufficient. Those are difficult to get right for API tokens.
BTW: This still works fine...