Proxmox team, go fuck yourselves.

Now I'm sure that I'll receive a lot of flack for this, but hear me out.

I've tried Proxmox and was quite pleased with its web UI. But I hate how much it locks me into their own little ecosystem.

I want to use btrfs on my drives. Why is this impossible, yet the hack that is ZoL is your obvious alternative? An alternative wherein I can't even compile and run my own kernel, because then ZoL suddenly fails? And don't you tell me to compile your stock config, when it's well over 15GB large in your source tree.

Proxmox is literally the MacOS of Linux distributions. Which was even more so made clear by me being called an idiot by possibly wanting to run Same on the PVE host. Because why on Earth would sysadmins want to?! Why on Earth would sysadmins be competent for wanting to?!!

You know what? I'll just convert those Proxmox servers to Arch and say fuck you to all the bells and whistles that's Proxmox' web UI. Because at least Arch allows me to make my own fucking choices, limited only by what's supported by the Linux operating system.

Perhaps Proxmox will consider btrfs stable in 2021. Because you know, despite it being stable today in 2018, Debian and Proxmox alike live 3 years in the past, i.e. 2015. I hate the Debian ecosystem because of that, but boy do I hate Proxmox even more so. Bloody fucking piece of shit it is!!! 😡

@Condor
If you look closely, you can read in the arch-wiki btrfs page, that there are still bugs, that cause data-loss.

Proxmox is perfectly fine software. Your concerns are with them not wanting to implement buggy software and a not-so-friendly community.

Again, nothing to do with Proxmox or the technology.

@FrodoSwaggins how long ago was this? I've been using btrfs for some 2-3 years now (on top of Arch so usually the recent kernels, which does a lot for btrfs stability), coming from ext4 because some fuckwad at the previous place I lived kept tripping the breakers, and btrfs seemed to be able to recover from the resulting server blackouts far better than ext4 could, probably due to its CoW nature.

As for issues with it, I couldn't find any, other than some oddly behaving WinBtrfs driver (in experimental state, so no surprise there). Granted, I've only used btrfs on single disks and RAID-0 and RAID-1 so far, not RAID-5 or RAID-6 which apparently still have issues. Not sure if using mdadm for the RAID array and btrfs on top of that would be a temporary solution, but my hunch is that it would be.. anyway, RAID 5 and 6 are the only big issues still acknowledged by the btrfs team.

@metamourge I've read the Arch Wiki page, but the closest to issues I could find was troubleshooting steps for when your btrfs array doesn't boot (for reasons that could range from blackouts to just not having included it in the kernel etc). The troubleshooting steps were things like adding btrfs to the initcpio.conf, then rebuilding initramfs so that you can use btrfs commands from there next time the system boots up. Basic troubleshooting essentially. But nothing related to bugs. Have you read the Arch Wiki page before posting this?

@xalys I agree that Proxmox is fine within its own ecosystem, just as much as Apple is. But the lack of ability to veer off-track and configure the system the way I want it to sure kills it for me. One could argue whether it's okay to run Samba on the host or use btrfs. But those things should be on the operator to decide in the context of the environment, not on the distribution maintainers.

@Condor
I think there is a red box somewhere, that says there is a data-loss bug, if u use it in RAID 5/6.

@metamourge my apologies, I've just checked in the browser and it looks to be the case indeed. For some reason the ArchWiki Viewer app seems to omit them. I should've checked more rigorously before drawing conclusions.

I'm pretty sure if you are the text-book definition of a prima donna.