245.4-2ubuntu1 throws stderr messages on timesyncd restart in its dhcp hook
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| chrony (Ubuntu) |
Fix Released
|
High
|
Unassigned | ||
| systemd (Ubuntu) |
Fix Released
|
Medium
|
Balint Reczey | ||
Bug Description
systemd has broken this by the split of timesyncd.
On a system that installed chrony (or other NTP servers) you'll have:
Desired=
| Status=
|/ Err?=(none)
||/ Name Version Architecture Description
+++-===
ii chrony 3.5-6ubuntu3 amd64 Versatile implementation of the Network Time Protocol
rc systemd-timesyncd 245.4-4ubuntu1 amd64 minimalistic service to synchronize local time with NTP servers
That left this behind:
/etc/
And if you now run a dhcp based time server push it will fail very badly.
P.S. This is triggered by the chrony autopkgtests but not an issue in chrony yet it is in systemd.
But due to that one can use [1] as easy reproducer, I wonder how systemd slipped by this issue in autopkgtest?
It contains:
systemctl try-restart systemd-
Which in the setup the system now has will trigger:
sudo dhclient dummy0
Failed to try-restart systemd-
P.S. I'll harden the chrony test to not stumble over this but it should be revisited for systemd to provide a better fix than causing this message every time (maybe pre-check if it is enabled and unmasked)?
Related branches
- Bryce Harrington (community): Approve
- Canonical Server: Pending requested
- Canonical Server packageset reviewers: Pending requested
-
Diff: 801 lines (+634/-5)10 files modifieddebian/README.container (+60/-0)
debian/changelog (+473/-0)
debian/chrony.conf (+18/-1)
debian/chrony.default (+4/-0)
debian/chrony.service (+2/-2)
debian/chronyd-starter.sh (+70/-0)
debian/control (+3/-1)
debian/docs (+1/-0)
debian/install (+1/-0)
debian/tests/control (+2/-1)
- Balint Reczey (community): Approve
- Canonical Server: Pending requested
- Canonical Server packageset reviewers: Pending requested
- git-ubuntu developers: Pending requested
-
Diff: 75 lines (+18/-17)3 files modifieddebian/changelog (+16/-0)
debian/postrm (+0/-16)
debian/tests/control (+2/-1)
| Christian Ehrhardt (paelzer) wrote | #1 |
| Changed in chrony (Ubuntu): | |
| status: | New → Fix Committed |
| Launchpad Janitor (janitor) wrote | #2 |
| Changed in chrony (Ubuntu): | |
| status: | Fix Committed → Fix Released |
| Changed in systemd (Ubuntu): | |
| assignee: | nobody → Balint Reczey (rbalint) |
| Balint Reczey (rbalint) wrote | #3 |
| Changed in systemd (Ubuntu Bionic): | |
| status: | New → Invalid |
| Changed in systemd (Ubuntu Eoan): | |
| status: | New → Invalid |
| Changed in systemd (Ubuntu): | |
| status: | New → In Progress |
| importance: | Undecided → Medium |
| summary: |
- 245.4-2ubuntu1 has broken dhcp based NTP updates + 245.4-2ubuntu1 has broken chrony autopkgtest |
| summary: |
- 245.4-2ubuntu1 has broken chrony autopkgtest + 245.4-2ubuntu1 throws stderr messages on timesyncd restart in its dhcp + hook |
| no longer affects: | chrony (Ubuntu Bionic) |
| no longer affects: | chrony (Ubuntu Eoan) |
| no longer affects: | systemd (Ubuntu Eoan) |
| no longer affects: | systemd (Ubuntu Bionic) |
| Changed in chrony (Ubuntu): | |
| importance: | Undecided → High |
| Launchpad Janitor (janitor) wrote | #4 |
| Changed in systemd (Ubuntu): | |
| status: | In Progress → Fix Released |
Chrony fix in focal completed tests and should migrate any minute.
But as that is only a mitigation of the symptoms not a fix to the problem in systemd-timesyncd at all.