Crash when running autoinstall with data generated from a previous install
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
subiquity |
New
|
Undecided
|
Unassigned |
Bug Description
Steps to reproduce:
1. Perform a subiquity installation with kvm-test.py (no autoinstall) and use the default value everywhere (next, next, next, next, done...)
$ scripts/kvm-test.py -m 2G --install --overwrite --iso jammy-live-
2. Once the installation has finished, copy over the /var/log/
3. Perform a second subiquity installation using the autoinstall file previously generated:
$ scripts/kvm-test.py -m 2G --install --overwrite --iso jammy-live-
The install crashes during the curtin partitioning stage.
I'm attaching the logs of the first and second installation. The autoinstall-
Reproduced today again using a kinetic ISO.
The invocations of `curtin install` in both runs look almost identical. Only the machine_id generated (and the resume_data directory) change ; which is expected.
The following differences are observed in curtin.log before everything starts failing:
https:/ /pastebin. ubuntu. com/p/cHgvR3V6C K/
* in red is where the autoinstall data was produced.
* in green is where the autoinstall data was consumed - and the install failed.