Red Hat Bugzilla – Bug 1249419
Please remove ConditionVirtualization=no from systemd-timesyncd.service
Last modified: 2015-08-10 05:53:00 EDT
Description of problem:
systemd-timesyncd can't be used to synchronise a guest VM's clock.
Version-Release number of selected component (if applicable):
Reproducible every time.
Steps to Reproduce:
1. grep ConditionVirtualization /usr/lib/systemd/system/systemd-timesyncd.service
Sadly you can't dictate that time synchronisation should not run just because the OS is running on virtualisation. It's true that the clock might be being synced some other way but you simply can't know and some guides will tell you to use NTP synchronisation in preference to hypervisor specific tools:
Always disabling it takes away the user's choice and leaves people baffled as to why things aren't working. I've got to admit that it would be useful if systemd logged when it abandoned a service due to failing a condition (e.g. "Not starting Time synchronisation because virtualisation=microsoft" etc).
This looks like it's going to be made specific to containers in later systemd releases: https://github.com/systemd/systemd/commit/4b16233e59b3edd8338819dad275cc94b5e5f6e7 .
OK, will backport the patch.
Note that you can always override the setting by using 'systemctl edit systemd-timesyncd' and adding '[Unit] ConditionVirtualization='.
Thanks for telling me about systemctl edit Zbigniew!
systemd-219-21.fc22 has been submitted as an update for Fedora 22.
* should fix your issue,
* was pushed to the Fedora 22 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing systemd-219-21.fc22'
as soon as you are able to.
Please go to the following url:
then log in and leave karma (feedback).
Thanks Zbigniew - this update resolves my problem.
systemd-219-21.fc22 has been pushed to the Fedora 22 stable repository. If problems still persist, please make note of it in this bug report.