unattended-updates failing in kinetic as AC power detection is bugged

[Gordon Lack]

The ConditionACPower testing code does not work correctly on some systems. This prevents the unattended-updates systemd service running as it thinks that a system (a NUC10i3FNYHN) is running on battery. It isn't - it just has a power supply that can *provide* power over USB.

This (unattended-updrades) was OK in Jammy, but started to fail when I upgraded to Kinetic.

(A systemd override file for apt-daily-upgrade.service to remove the condition gets the service working again.)

The problem is similar (or perhaps identical) to the problem with the on_ac_power script (from the powermgmt-base package).

That has been reported in bug #1980991. There are more details there.

[Gordon Lack (gordon-lack)]

This also affects various other services.

apt-daily
e2scrub_all
logrotate
man-db
plocate-updatedb

[Gordon Lack (gordon-lack)]

This has (apparently) been fixed by https://github.com/systemd/systemd/pull/24220

[Nick Rosbrook (enr0n)]

I confirmed this patch is not present in kinetic, but it will be in lunar.

[Nick Rosbrook (enr0n)]

This fell off the radar for Kinetic, and at this point I don't think we will do another systemd SRU for Kinetic.

[Nick Rosbrook (enr0n)]

The referenced commit should be in Lunar and newer. Marking Won't Fix for Kinetic as we are not planning on another SRU for that release.

[Gordon Lack (gordon-lack)]

I've tested /lib/systemd/systemd-ac-power on my laptop (which gives 0/1 for mains/battery) and it now returns 0 (mains) for my NUC10i3FNYHN.

(But the fix isn't in Lunar for on_ac_power.)

[Utkarsh Gupta (utkarsh)]

Ubuntu 22.10 (Kinetic Kudu) has reached end of life, so this bug will not be fixed for that specific release.

[Gordon Lack (gordon-lack)]

All of which should be irrelevant, as it is still an issue in Lunar *and* Jammy (the latest LTS).

What is the point in reporting a bug against a non-LTS version if the response it to do nothing then close it 6-momnths later when that release goes out of support, without any one checking whether it is still an issue on the latest release?

NOTE: that this is still an issue on the current LTS release (Jammy) *and* on Lunar. Just closing this report does not make the bug go away.

[Nick Rosbrook (enr0n)]

I thought this (the CondiationAcPower check) was never a problem in Jammy? I was under the impression that the systemd part of this was only a problem in Kinetic. Or are you referring to the on_ac_power part, and therefore more appropriately about bug 1980991?

[Theodore Ts'o (tytso)]

This is not a bug for e2fsprogs, so marking it invalid (for e2fsprogs) so it stops spamming my inbox.

[Gordon Lack (gordon-lack)]

=> Nick Rosbrook (enr0n)

Sorry - you're correct. This one is fixed (I removed my override workarounds on June 12, so it must be...).

Just a pity that the report is left as "New" and "Won't fix", rather than "Fixed in ...".

[Nick Rosbrook (enr0n)]

Since the change was in systemd, the appropriate resolution for the other packages is actually "Invalid." It was not fixed in Kinetic, so "Won't Fix" is appropriate there.