Seems a serious issue with silverblue update boot stuck - a start job is running for /dev/disk/by-uuid/8d21faa4-0cac-45b9-9d86-2b54e2fe1f02

Ask Fedora Ask in English
,
1

seems a serious issue with latest silverblue updates boot stuck - a start job is running for /dev/disk/by-uuid

I believe its related to the newest kernel, i am running silverblue 35, fine on kernel 5.16.20-200.fc35.x86_64, but a recent update including the following repos Enabled rpm-md repositories: fedora-35-updates fedora-35 vivaldi rpmfusion-free rpmfusion-nonfree rpmfusion-nonfree-updates rpmfusion-free-updates slack wants

kernel-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-core-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-devel-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-headers-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-modules-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-modules-extra-5.17.4-200.fc35.x86_64 (fedora-35-updates)
kernel-srpm-macros-1.0-6.fc35.noarch (fedora-35)

and a syctl reboot after update fails for continue past a start job is running for /dev/disk/by-uuid/8d21faa4-0cac-45b9-9d86-2b54e2fe1f02

so how to debug.

2

It looks like it’s having trouble mounting a disk. If you have more than one disk, it’s possible that one of them might not be healthy.

1 Like
3

kernel 5.16 boots fine on it, btrfs enabled triple 2TB NVME drives, and their new. doubtful its a faulty drive.

dmesg | grep nvme
[    1.726231] nvme nvme0: pci function 0000:03:00.0
[    1.726427] nvme nvme1: pci function 10000:01:00.0
[    1.726442] nvme 10000:01:00.0: PCI INT A: not connected
[    1.726532] nvme nvme2: pci function 10000:02:00.0
[    1.726547] nvme 10000:02:00.0: PCI INT A: not connected
[    1.732761] nvme nvme0: missing or invalid SUBNQN field.
[    1.732783] nvme nvme0: Shutdown timeout set to 8 seconds
[    1.732783] nvme nvme1: missing or invalid SUBNQN field.
[    1.732793] nvme nvme2: missing or invalid SUBNQN field.
[    1.732798] nvme nvme1: Shutdown timeout set to 8 seconds
[    1.732810] nvme nvme2: Shutdown timeout set to 8 seconds
[    1.759557] nvme nvme0: 32/0/0 default/read/poll queues
[    1.759791] nvme nvme1: 32/0/0 default/read/poll queues
[    1.759902] nvme nvme2: 32/0/0 default/read/poll queues
[    1.762881]  nvme0n1: p1 p2
[    1.763281]  nvme2n1: p1
[    1.763349]  nvme1n1: p1
[    1.770851] BTRFS: device label fedora_fedora devid 2 transid 28418 /dev/nvme1n1p1 scanned by systemd-udevd (728)
[    1.771226] BTRFS: device label fedora_fedora devid 3 transid 28418 /dev/nvme2n1p1 scanned by systemd-udevd (773)
[    1.772391] BTRFS: device label fedora_fedora devid 1 transid 28418 /dev/nvme0n1p2 scanned by systemd-udevd (780)
[    3.293406] BTRFS info (device nvme0n1p2): flagging fs with big metadata feature
[    3.293410] BTRFS info (device nvme0n1p2): disk space caching is enabled
[    3.293411] BTRFS info (device nvme0n1p2): has skinny extents
[    3.338887] BTRFS info (device nvme0n1p2): enabling ssd optimizations
[    3.880891] BTRFS info (device nvme0n1p2): use zstd compression, level 1
[    3.880895] BTRFS info (device nvme0n1p2): disk space caching is enabled
[    6.209717] EXT4-fs (nvme0n1p1): mounted filesystem with ordered data mode. Opts: (null). Quota mode: none.
[    6.280648] BTRFS info (device nvme0n1p2): disk space caching is enabled
[    6.280846] BTRFS info (device nvme0n1p2): disk space caching is enabled
2 Likes
4

bump… nobody has any other ideas?

5

Hi, I am sorry that I can not provide any insightful information besides saying that this is also affecting me.

I can boot no problem in Silverblue 35, but Silverblue 36 gives me this problem more often than not. To the point that I have not been able to finish booting in order to obtain a previous boot log, so I can only attach a poor picture:

43ff4ee35fd62712cdcb45113aa6caf98ba2b017.png
43ff4ee35fd62712cdcb45113aa6caf98ba2b017.png1847×269 250 KB

rpm-ostree deployment information:

  fedora:fedora/36/x86_64/silverblue
                   Version: 36.20220511.0 (2022-05-11T00:48:12Z)
                BaseCommit: 5c70836453ffbd07757cabeb4c1de5389b95d45d7ec6fe8d2397084e1587fcd7
              GPGSignature: Valid signature by 53DED2CB922D8B8D9E63FD18999F7CBF38AB71F4
       RemovedBasePackages: firefox 100.0-2.fc36
           LayeredPackages: akmod-nvidia xorg-x11-drv-nvidia
6

okay welp, that says at least im not the only one. are you using multiple drives i have 3x2TB ssd configured, and yes same F35 boots fine. F36, and Silverblue 36 same result.

7

Yes, indeed: I have an HDD (1TB), an SSD (250GB) and a NVME (500GB).

My btrfs volume spans partitions on both the NVME and the SSD. There are other partitions but are not mounted on boot.

8

As an update to this, I must say that it has ceased to happen. Either I am being extremely lucky, or some update may have been fixed the issue.

1 Like