vorta-0.8.12-2 failing autopkgtests with python-secretstorage as trigger

Bug #2045997 reported by Chris Peterson
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
python-secretstorage (Ubuntu)
Fix Released
Undecided
Unassigned
vorta (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

autopkgtest log:

https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/amd64/v/vorta/20231202_011527_de8c9@/log.gz

The issue appears to be:

> Message: 08:40:57.525: couldn't access control socket: /tmp/autopkgtest.ifpw2s/autopkgtest_tmp/vorta/run/keyring/control: No such file or directory

which results in the test timing out.

I can't reproduce this locally, building in a noble (w/ -proposed) schroot and running autopkgtest in a noble (no -proposed) schroot.

Additionally, the other other autopkgtests for vorta-0.8.12-2 don't have the same issue. This only happens when triggered by python-secretstorage/3.3.3-2

Revision history for this message
Chris Peterson (cpete) wrote :

~liushuyu-011 confirmed for me he is able to reproduce the autopkgtest error with a qemu runner. There must be a difference between the autopkgtest image and my sbuild/schroot setup.

Revision history for this message
Chris Peterson (cpete) wrote :

I can confirm this failure with a local qemu autopkgtest server. From the following lines in the log:

.dbus-daemon[2104]: [session uid=1000 pid=2104] Activating service name='org.freedesktop.secrets' requested by ':1.1' (uid=1000 pid=2105 comm="python3.11 -m pytest tests" label="unconfined")
378s ** Message: 22:28:34.588: couldn't access control socket: /tmp/autopkgtest.SbzXU7/autopkgtest_tmp/vorta/run/keyring/control: No such file or directory
378s discover_other_daemon: 0dbus-daemon[2104]: [session uid=1000 pid=2104] Successfully activated service 'org.freedesktop.secrets'
378s dbus-daemon[2104]: [session uid=1000 pid=2104] Activating service name='org.gnome.keyring.SystemPrompter' requested by ':1.2' (uid=1000 pid=2110 comm="/usr/bin/gnome-keyring-daemon --start --foreground" label="unconfined")
378s dbus-daemon[2104]: [session uid=1000 pid=2104] Activating service name='org.a11y.Bus' requested by ':1.3' (uid=1000 pid=2117 comm="/usr/libexec/gcr-prompter" label="unconfined")
378s dbus-daemon[2104]: [session uid=1000 pid=2104] Successfully activated service 'org.a11y.Bus'
378s dbus-daemon[2125]: Activating service name='org.a11y.atspi.Registry' requested by ':1.0' (uid=1000 pid=2117 comm="/usr/libexec/gcr-prompter" label="unconfined")
378s dbus-daemon[2104]: [session uid=1000 pid=2104] Successfully activated service 'org.gnome.keyring.SystemPrompter'
378s dbus-daemon[2125]: Successfully activated service 'org.a11y.atspi.Registry'
378s SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry

my guess so far is that it's launching a new gnome-keyring-daemon in the test's dbus session and blocking the tests from continuing (--foreground instead of --daemonize).

This is sort of expected, since python-secretstorage now pulls in gnome-keyring on install where it hadn't before. Although I'm not exactly sure how this request takes over the pytest process.

I've yet to get gnome-keyring-daemon to start in the test's dbus session and not block the tests from running, which I think is the proper solution.

Otherwise, if the tests are valid without an external secret service provider (like gnome-keyring), then we should patch the tests to use whatever fallback it had been using before when python-secretstorage didn't find a provider for the org.freedesktop.secrets service.

tags: added: update-excuse
Revision history for this message
Dmitry Shachnev (mitya57) wrote :

In python-secretstorage 3.3.3-3 I had to demote gnome-keyring to Suggests (see https://lists.debian.org/debian-devel/2024/01/msg00098.html), so now vorta autopkgtest passes again.

Changed in python-secretstorage (Ubuntu):
status: New → Fix Released
Changed in vorta (Ubuntu):
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.