| 18 Aug 2024 |
 Whovian9369 | In reply to @k900:0upti.me
Check journalctl on the NixOS side, are there any errors? Aug 18 16:50:18 nixos-wsl kernel: FS-Cache: Duplicate cookie detected then its information
Aug 18 16:50:18 nixos-wsl unknown: /usr/lib/wsl/lib/libcuda.so.1 is not a symbolic link
Aug 18 16:50:18 nixos-wsl unknown: WSL (2) ERROR: WaitForBootProcess:3352: /sbin/init failed to start within 10000ms
Aug 18 16:50:18 nixos-wsl systemd-journald[79]: Received client request to flush runtime journal.
Aug 18 16:50:18 nixos-wsl systemd-journald[79]: File /var/log/journal/3f0405037402365e4122e654618a2843/system.journal corrupted or uncleanly shut down, renaming and replacing.
...
Aug 18 16:50:18 nixos-wsl systemd-sysctl[263]: Couldn't write '16' to 'kernel/sysrq', ignoring: No such file or directory
These are what stood out to me, and mostly repeat a few times
| 20:56:28 |
 K900 | Yeah that's fine | 20:56:41 |
| Whovian9369 | In reply to @k900:0upti.me
Check journalctl on the NixOS side, are there any errors? * Aug 18 16:50:18 nixos-wsl kernel: FS-Cache: Duplicate cookie detected then its information
Aug 18 16:50:18 nixos-wsl unknown: /usr/lib/wsl/lib/libcuda.so.1 is not a symbolic link
Aug 18 16:50:18 nixos-wsl unknown: WSL (2) ERROR: WaitForBootProcess:3352: /sbin/init failed to start within 10000ms
...
Aug 18 16:50:18 nixos-wsl systemd-journald[79]: Received client request to flush runtime journal.
Aug 18 16:50:18 nixos-wsl systemd-journald[79]: File /var/log/journal/3f0405037402365e4122e654618a2843/system.journal corrupted or uncleanly shut down, renaming and replacing.
...
Aug 18 16:50:18 nixos-wsl systemd-sysctl[263]: Couldn't write '16' to 'kernel/sysrq', ignoring: No such file or directory
These are what stood out to me, and mostly repeat a few times
| 20:56:41 |
| K900 | As in none of that is fatal | 20:56:51 |
| K900 | Though init not starting is weird | 20:56:55 |
| K900 | Is this NixOS 24.05 or unstable? | 20:57:02 |
| Whovian9369 | Should be Unstable | 20:57:10 |
| K900 | Oh | 20:57:19 |
| K900 | And did you update in between restarting the VMs? | 20:57:28 |
| Whovian9369 | You mean while we were experimenting? No, not to my knowledge | 20:57:50 |
| K900 | Do you have
[wsl2]
kernelCommandLine = cgroup_no_v1=all
in your .wslconfig? | 20:58:02 |
| Whovian9369 | I don't have anything in my C:\Users\Whovian\.wslconfig (Everything is commented out.) | 20:59:05 |
| K900 | Yeah, you need to set that | 20:59:33 |
| K900 | For systemd 256 to start correctly | 20:59:36 |
| K900 | Otherwise things get Funny | 20:59:40 |
| Whovian9369 | Ah damn it, thanks 😅
I'll try that in a moment | 20:59:49 |
| Whovian9369 | wsl.exe --shutdown, add that, and try again? | 21:00:00 |
| K900 | Thank Microsoft for not shipping a cgv2-by-default kernel like three months after Ubuntu shipped it as a requirement | 21:00:16 |
| K900 | (and like multiple years after everyone else decided it was a good idea) | 21:00:28 |
| Whovian9369 | â•â”€whovian@nixos-wsl ~
╰─$ sudo whoami
root
Okay cool, this is a good sign
| 21:02:34 |
| Whovian9369 | And the errors on every tab are magically gone, which means that all of the services are starting as I'd expect again.
Thank you so so much for the help and time, K900. | 21:03:12 |
| K900 | I just realized we can probably get better errors if we actually dup systemd logs to kmsg | 21:04:57 |
| K900 | CC @nzbr:nzbr.de | 21:05:02 |
| K900 | Oh never mind | 21:10:06 |
| K900 | systemd is trying to be smart | 21:10:10 |
| K900 | And write to /dev/console | 21:10:18 |
| K900 | And guess what WSL does to that | 21:10:23 |
| Whovian9369 | Then as a followup question, do you know if there's a way to automatically check for the .wslconfig entry (without reading the file directly) to warn a user to add it? | 21:11:49 |
| K900 | No | 21:11:58 |
| K900 | We basically have no user facing communication until after systemd | 21:12:13 |
