Bugzilla – Bug 1224824
Install fails if a previous installation has been aborted during migration backup
Last modified: 2024-05-23 20:04:54 UTC
User-Agent: Mozilla/5.0 (Windows NT 10.0; rv:126.0) Gecko/20100101 Firefox/126.0 Build Identifier: Installation on a Aeon RC1 was started normal. The installations was terminated by the user during the backup of home (because it took longer than expected and the user had to terminate the installation). Subsequent installations failed, even with newly flashed installations devices. The reason seems to be that a snapshot of /home on the source installation already exists and a creating a new snapshot fails. Reproducible: Always Steps to Reproduce: 1. Start installation on a system with Aeon installed 2. Select to migrate existing users 3. Interrupt the installation during the backup 4. All subsequent installations with migration fails Actual Results: Installation with migration not possible anymore Expected Results: Previous snapshots (/home/homebk) should be deleted before a new snapshot is created
Created attachment 875008 [details] logfile
This log doesn't look valid, so I'd like to understand what you did that you didn't document in your report first /dev/disk/by-id/dm-name-cr_root doesn't seem like a valid device path I'd expect the get_disks function to detect Can you please explain your hardware/disk setup here? I don't want to prioritise bugs on artificial scenarios/not-real hardware :)
Even with the mysterious disk..addressed in https://github.com/sysrich/tik/pull/15
Fixed in tik 1.0.0
My Aeon RC1 is installed on a LUKS encrypted drive. After startup with the installer USB, I opened a terminal to open the LUKS drive. Then started the installer as normal. The migration runs through like normal (if there wasn't a previous terminated migration as described in this bug report).
That’s fine, what you did manually tik now does itself as a feature Sweet