about summary refs log tree commit diff
path: root/content/blog/nix-raid-systemd-boot.md
diff options
context:
space:
mode:
authorFranck Cuny <franck@fcuny.net>2024-07-03 16:32:51 -0700
committerFranck Cuny <franck@fcuny.net>2024-07-03 16:32:51 -0700
commitd481251d4148a9e90cf71aa1c11a8f8e077336a4 (patch)
treea5dd8993718db7cfc7b13ef2a24a3fa2e15b582f /content/blog/nix-raid-systemd-boot.md
parentsimplify the layout (diff)
downloadfcuny.net-d481251d4148a9e90cf71aa1c11a8f8e077336a4.tar.gz
some more cleanup
Diffstat (limited to 'content/blog/nix-raid-systemd-boot.md')
-rw-r--r--content/blog/nix-raid-systemd-boot.md7
1 files changed, 5 insertions, 2 deletions
diff --git a/content/blog/nix-raid-systemd-boot.md b/content/blog/nix-raid-systemd-boot.md
index b31d29c..cd020f2 100644
--- a/content/blog/nix-raid-systemd-boot.md
+++ b/content/blog/nix-raid-systemd-boot.md
@@ -1,13 +1,14 @@
 ---
 title: Workaround md raid boot issue in NixOS 22.11
 date: 2023-01-10
-tags:
-- nixos
 ---
+
 For about a year now I've been running [NixOS](https://nixos.org/ "NixOS") on my personal machines. Yesterday I decided to go ahead and upgrade my NAS from NixOS 22.05 to [22.11](https://nixos.org/blog/announcements.html#nixos-22.11). On that machine, all the disks are encrypted, and there are two RAID0 devices. To unlock the drives, I log into the [SSH daemon running in `initrd`](https://nixos.wiki/wiki/Remote_LUKS_Unlocking), where I can type my passphrase. This time however, instead of a prompt to unlock the disk, I see the following message:
+
 ```
 waiting for device /dev/disk/by-uuid/66c58a92-45fe-4b03-9be0-214ff67c177c to appear...
 ```
+
 followed by a timeout and then I'm asked if I want to reboot the machine. I do reboot the machine, and same thing happens.
 
 Now, and this is something really great about NixOS, I can boot to the previous generation (on 22.05), and this time I'm prompted for my password, the disks are unlocked, and I can log into my machine. This eliminates the possibility of a hardware failure! I also have a way to get a working machine to do more build if needed. Knowing that I can easily switch from a broken generation to a working one gives me more confidence in making changes to my system.
@@ -17,10 +18,12 @@ I then reboot again in the broken build, and drop into a `busybox` shell. I look
 My laptop has a similar setup, but without RAID devices. I had already updated to 22.11, and had rebooted the laptop without issues. To be sure, I ran another update and rebooted, and I was able to unlock the drive and log into the machine without problem.
 
 From here I have enough information to start searching for an issue similar to this. I got pretty lucky and two issues I found were:
+
 - [Since systemd-251.3 mdadm doesn't start at boot time #196800 ](https://github.com/nixoS/nixpkgs/issues/196800)
 - [Won't boot when root on raid0 with boot.initrd.systemd=true #199551 ](https://github.com/nixoS/nixpkgs/issues/199551)
 
 The proposed solution was easy:
+
 ```diff
 @@ -43,7 +43,7 @@
    };