Subiquity sometimes fails to do the post-install reboot on focal/amd64
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
subiquity |
Confirmed
|
Undecided
|
Unassigned |
Bug Description
Starting with the 2021-02-01 ubuntu-
Reproducer via preseeded install (answers.yaml):
1. mkdir /tmp/isofail ; cd /tmp/isofail
2. download [3] and [4]
3. download the latest focal-live-
4. qemu-img create -f qcow2 disk0.qcow2 8G
5. virsh create fail.xml
Try this a few times and you should hit the reboot hang. Note: the VM gets undefined at shutdown so on "success" (bug not reproduced) it will just disappear. Just try the same command again.
This happens with a host system on Bionic, Bionic + cloud archive, Hirsute.
I could *not* reproduce it with a bare kvm/qemu install like:
kvm -m 2048 -boot d -cdrom ./focal-
This command results in an install loop (the cdrom is never removed) that never got stuck in my testing.
[1] https:/
[2] https:/
[3] https:/
[4] https:/
tags: | added: fr-1107 |
This bug has been reported on the Ubuntu ISO testing tracker.
A list of all reports related to this bug can be found here: iso.qa. ubuntu. com/qatracker/ reports/ bugs/1914285
http://