| 20 Jan 2025 |
lassulus | like create should be idempotent | 21:18:13 |
lassulus | but please have backups and proof read the created script :) | 21:18:26 |
guiniol | and I wanted to create the datasets as disko would create them because at the very end, I will wipe the machine and reinstall everything from backups | 21:18:27 |
magic_rb | Ah | 21:18:37 |
magic_rb | Dry run then | 21:18:43 |
guiniol | so I don't want to have an incompatible option somewhere taht would make the zfs receive fail | 21:18:59 |
guiniol | so, I did try --dry-run but apparently my flake on its own is not enough. I create the system with nixos-anywhere and disko complains about You must set the option ‘boot.loader.grub.devices’ or 'boot.loader.grub.mirroredBoots' to make the system bootable. | 21:19:48 |
guiniol | I think there was something about a bootable device in the help, so let me try that again | 21:20:13 |
guiniol | alright, I think I have it | 21:34:14 |
guiniol | and there are basically no options being added :D | 21:34:29 |
guiniol | Alright, another question: I had to disable the zfs-mount service because the ZFS dataset are also being mounted by autogenerated systemd units. Did I do something wrong? I am using non-legacy mount type datasets. | 22:38:58 |
| 22 Jan 2025 |
| Ben joined the room. | 15:14:02 |
| 23 Jan 2025 |
| @asymmetric:matrix.dapp.org.uk joined the room. | 14:58:41 |
@asymmetric:matrix.dapp.org.uk | hey, i just tried to set-up a mdadm raid0 and got this error:
+ sgdisk --zap-all /dev/md/raid0
Caution: invalid main GPT header, but valid backup; regenerating main header
from backup!
Error! Couldn't seek to partition table!
Caution! After loading partitions, the CRC doesn't check out!
Warning! One or more CRCs don't match. You should repair the disk!
Main header: ERROR
Backup header: OK
Main partition table: ERROR
Backup partition table: ERROR
Invalid partition data!
GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.
the disk-config is here: https://gist.github.com/asymmetric/6e12e6da1b1d7135354476174c37c447
| 14:59:46 |
@asymmetric:matrix.dapp.org.uk | is this a bug? | 14:59:50 |
lassulus | hmm, seems like the zap is not powerful enough :D | 15:55:44 |
lassulus | I guess we can ignore the return value for zap-all in that case? maybe use wipefs or so instead | 15:56:50 |
| 24 Jan 2025 |
phaer | should sgdisk --zap-all be called at all for mdadm devices? Not an mdadm user myself anymore, but my impression was that you'd do a mdadm -S and then zap/wipe the underlying physical disk? | 11:32:33 |
| mutantmell joined the room. | 17:55:47 |
| mutantmell set a profile picture. | 18:13:49 |
| mutantmell changed their profile picture. | 18:14:00 |
| 25 Jan 2025 |
bumperboat (UTC+2) | i added a btrfs subvol to my disko config, it doesnt seem to be created after the fact - is it expected to? | 07:29:35 |
srestegosaurio (on tchncs.de) | In reply to @bumperboat:matrix.org i added a btrfs subvol to my disko config, it doesnt seem to be created after the fact - is it expected to? I was having similar experience. I ended up creating the volume myself. | 11:21:58 |
srestegosaurio (on tchncs.de) | Disco would take care of mounting it for me tho. | 11:22:14 |
bumperboat (UTC+2) | ok thx, i see it does make the mount unit, and fails because it didnt make the subvol | 11:23:18 |
srestegosaurio (on tchncs.de) | In reply to @bumperboat:matrix.org ok thx, i see it does make the mount unit, and fails because it didnt make the subvol Yep, exactly the same. | 11:29:40 |
bumperboat (UTC+2) | thx for the confirmation | 11:43:54 |
lassulus | support for running the createscript after rebuild again is very new and not thoroughly tested. You have to run the create script (not the diskoScript) manually and it should create new subvolumes, lvs and datasets. But better have backups :) | 11:48:18 |
srestegosaurio (on tchncs.de) | In reply to @lassulus:lassul.us support for running the createscript after rebuild again is very new and not thoroughly tested. You have to run the create script (not the diskoScript) manually and it should create new subvolumes, lvs and datasets. But better have backups :) Didn't know about this, thanks for the heads up. | 12:09:56 |
arjan5 | Question: is it possible to define a disk/partition optionally? I want the boot process to continue if a certain disk is missing. | 19:38:39 |