| 2 Jul 2024 |
 lassulus | hmm, this works only in the case where you would have the current existing structure defined in nix which lives on your disk? | 06:42:49 |
| lassulus | not a very common usecase I would say | 06:42:55 |
| lassulus | or you think just for the disk? | 06:43:08 |
 colemickens | I'm not sure I followed - to recap, for first provision, I want to destroy/format, and then presumably update my disk-config.nix to say "okay, no longer run destroy on disk2". Since disk2 is going to persist as it's a hetzner volume. | 06:44:14 |
| colemickens | disk1 is just the ephemeral vm disk, so I always need it destroyed | 06:44:30 |
| colemickens | It might not be super duper common, but I feel like it's an interesting use-case. maybe :). | 06:44:56 |
| colemickens | I guess.. I mean, I could just write my own destroy command and prepend it to my overriden diskoScript.. | 06:45:50 |
| colemickens | nice, that worked. wipefs'd + dd'd off the first 512M. seems to have left my /home contents. | 06:53:26 |
| lassulus | you can run the disko internal destroy command and just run it on the disks you want | 06:53:36 |
| colemickens | I saw jq and bailed :) but yeah, I probably should just have done that | 06:54:04 |
| colemickens | for now this works, gotta figure out why this isn't playing nice with the arm64 version now. | 06:54:20 |
| 3 Jul 2024 |
 matthewcroughan | If vendor kernels are in use, it's easy to make the disk image builder stall | 18:24:34 |
| matthewcroughan | blikvm-v1-disko-images> UEFI firmware (version built at 00:00:00 on Jan 1 1980)
blikvm-v1-disko-images> EFI stub: Decompressing Linux Kernel...
blikvm-v1-disko-images> EFI stub: Loaded initrd from LINUX_EFI_INITRD_MEDIA_GUID device path
blikvm-v1-disko-images> EFI stub: Generating empty DTB
blikvm-v1-disko-images> EFI stub: Exiting boot services...
| 18:24:36 |
| matthewcroughan | https://github.com/nix-community/disko/commit/bb75ad0b1d5a63884586517920b9a189721fc54d | 18:28:34 |
| matthewcroughan | This commit was so much more problematic than I anticipated | 18:28:42 |
| matthewcroughan | I should make it possible to override this | 18:28:50 |
| 4 Jul 2024 |
|  eyJhb left the room. | 10:58:18 |
