Trying to understand my deployments and fix my own mistake

General
1

Hello,
I’ve done a fresh Bazzite install and I am very impressed by this distro. Coming from Mint, I have lots to learn though.

I automounted my internal HDD as one of the first steps. Apparently I made some mistake, because after the reboot, I got to the login screen, entered my password, then the screen went black for a second and I came back to the login screen again. Tried 3 times, same result. So I did restart and used the second deployment which booted normally. I fixed my mistake with the automount and pinned my current deployment. So far so good.

But everytime I boot the PC now, I’m going through this cycle again. Boot - login screen - have to restart - have to select second deployment - boot.

rpm-ostree status
State: idle
Deployments:
  ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-nvidia:stable
                   Digest: sha256:60fc301075672049b99767c49013ed14dc4b4d06896ed492d29984535539de62
                  Version: 40.20240912.0 (2024-09-13T05:53:37Z)
                     Diff: 54 upgraded

● ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-nvidia:stable
                   Digest: sha256:019b85f3ba44fcd0f7304eafe864ca9f4995da26fabadd70a0a5441de00fb695
                  Version: 40.20240908.0 (2024-09-09T05:03:59Z)
                   Pinned: yes

If I read this status correctly, my current deployment is the one I pinned and that is working. But why do I always go back to the other deployment? And also, it says “54 upgraded” - so I suppose, I should prefer the first deployment over the second?

I’m sorry if this is a very dumb question, but I read through the documentation and still cannot fix this. Please let me know if any more information is needed. Thank you.

2

I tried to rebase to the pinned deployment but it’s giving me an error:

bazzite-rollback-helper rebase 40-stable-20240908
Rebasing to ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-nvidia:40-stable-20240908. Continue? [y/N]:y
rebasing to bazzite-nvidia:40-stable-20240908
Pulling manifest: ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-nvidia:40-stable-20240908
error: Creating importer: Failed to invoke skopeo proxy method OpenImage: remote error: reading manifest 40-stable-20240908 in ghcr.io/ublue-os/bazzite-nvidia: manifest unknown

Searching the error pointed me to the signing error that happened in July so I did:

curl -sL https://fix.universal-blue.org/ | sudo bash 
[sudo] password for ******: 
Fetching the new public key from ublue-os's github repo...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   178  100   178    0     0    878      0 --:--:-- --:--:-- --:--:--   881
Updating the path to the public key in the container policy...
Updating the system...
Pulling manifest: ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-nvidia:stable
No upgrade available.

But I’m still getting the same error. How can I fix it, please?

3

I have found the solution in the answer overflow:

rpm-ostree cleanup -p

Thank you for providing this resource, I’m leaving the answer here in case anyone else runs into a similar situation.

1 Like