** Changed in: systemd (Ubuntu)
Status: New => Won't Fix
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubuntu Core 18
To mana
Fwiw, it looks like the latest core18 edge build that includes
https://launchpad.net/ubuntu/+source/cloud-init/21.4-0ubuntu1~18.04.1
make the problem go away. It seems https://bugs.launchpad.net/cloud-
init/+bug/1946003 is what caused it.
The old code had a udev rule that would call into the cloud
> It's not something that snapd writes out, it's curious where this
comes from
it's not written out, it's internally handled/added by pid1, and it
*should* be there, because you do want (for example) the mount to be
stopped if its backing device is removed. However when the mount is
stopped, the B
I attached the /run/cloud-init and /var/log/cloud-init logs of the
good/bad run - I looked over the diff via "diff -u <(cut -f 4- -d:
/tmp/cloud-init-good-core18-r2206/cloud-init.log) <(cut -f 4- -d:
/tmp/cloud-init-bad-core18-r2208/cloud-init.log)" but couldn't see
anything standing out there (but
** Attachment added: "cloud-init logs for the "good" run"
https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1949089/+attachment/5542024/+files/cloud-init-good-core18-r2206.tar.gz
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Attachment added: "cloud-init logs for the "bad" run"
https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1949089/+attachment/5542025/+files/cloud-init-bad-core18-r2208.tar.gz
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
h
I spend a bit more time bisecting the various core18 snaps to see at
what revision the failure started to become reproducible. Interestingly
I found that it started with r2208 AFAICT. It's all a bit annoying
because it's a race so I don't trust it 100% but the previous r2206 got
2 good runs in a ro
@Dan Thanks, this find about BindsTo is super interesting! It's not
something that snapd writes out, it's curious where this comes from.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
This added BindsTo is what's causing the problem:
ddstreet@localhost:~$ systemctl show -p BindsTo snap-hello\\x2dworld-29.mount
BindsTo=
ddstreet@localhost:~$ sudo systemctl daemon-reload
ddstreet@localhost:~$ systemctl show -p BindsTo snap-hello\\x2dworld-29.mount
BindsTo=dev-loop6.device
ubu
Sorry to clarify that, the dep is added to the mount over a daemon-
reload, but only on core:
ddstreet@localhost:~$ systemctl list-dependencies snap-hello\\x2dworld-29.mount
snap-hello\x2dworld-29.mount
● ├─-.mount
● ├─snap.mount
● ├─system.slice
● └─var-lib-snapd.mount
ddstreet@localhost:~$ syst
Note one interesting point is that in a core18 system, a squashfs mount
(any loop-based mount, really) includes a dep from the mount to its loop
device, while the non-core system does not:
ddstreet@localhost:~$ cat /etc/issue
Ubuntu Core 18 on \4 (\l)
ddstreet@localhost:~$ systemctl list-dependen
Thanks ddstreet, I agree 918e6f1c0151429f5095355f4f3f74f16e79724a could
be related as well, but unfortunately that commit produced quite some
fall-out.
Nevertheless, I've prepared a new systemd build in my PPA and the
"core18_2028_amd64.snap" at https://people.ubuntu.com/~slyon/uc18/
cherry-pi
> @slyon you might want to see if commit
918e6f1c0151429f5095355f4f3f74f16e79724a fixes this
(and there are a couple follow on commits to this)
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
T
I spend a bit of quality time with this bug today and it seems there is (also?)
a kernel dimension to it. I ran the following:
"""
$ git status
On branch tests-use-core18-from-gce
Your branch is up to date with 'sergiocazzolato/tests-use-core18-from-gce'.
$ git diff
diff --git a/overlord/devicesta
@slyon you might want to see if commit
918e6f1c0151429f5095355f4f3f74f16e79724a fixes this
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units
Thanks Lukas! It's very likely that there are several issues in this
area, because while I can confirm that your latest image fixes the
reproducer I pasted before (in the Google cloud, too), the spread test
failure is still reproducible with the "tests/main/services-disabled-
kept-happy" test.
I n
Thank you, having a reliable reproducer is very helpful! I can confirm
this reproduces locally for me on QEMU.
I prepared a new systemd package [0], containing backports from the
above mentioned PR that indeed sounds very related (I wonder why the
issue didn't happen before, tho, as this is unrela
One more update: I verified that the steps from comment #17 can be used
to reproduce the issue in an Ubuntu Core image started in qemu, and
without using the spread. I verified it with the 20211105 image provided
by Lukas in https://people.ubuntu.com/~slyon/uc18/
To reproduce it, start the image i
Might be related to https://github.com/systemd/systemd/issues/11342
** Bug watch added: github.com/systemd/systemd/issues #11342
https://github.com/systemd/systemd/issues/11342
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https:
An update to the comment above. We have found out a sequence of commands
that seems to reproduce the issue reliably: in the spread session
started by the command from the comment above, if you type this:
systemctl stop 'snap-disabled\x2dsvcs\x2dkept-x1.mount'
systemctl daemon-reload
#
> can you provide steps on how this is reproducable then?
So far we have not been able to isolate the problem, so the only way so
far is to run the spread tests from the snapd directory (from the master
branch):
spread -debug -repeat 20 google:ubuntu-
core-18-64:tests/main/services-disabled-k
> and in QEMU it seems that this bug is not reproducible
can you provide steps on how this is reproducable then?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails
I'm coming with bad news. Today both me and Michael have individually
retested both snaps you provided (20211102 and 20211105), and the bug
happened with both of them. It's possible that yesterday Michael
accidentally used QEMU for his tests, without realizing (and in QEMU it
seems that this bug is
** No longer affects: systemd (Ubuntu Jammy)
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubuntu Core 18
To manage notifications ab
About the question about Focal/hirsute/uc20 - we have not observed the
issue there.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubu
That's certainly possible. In the first attempt (that failed according
to mvo's tests) I only backported
https://github.com/systemd/systemd-stable/commit/04eb582acc203eab0bc5c2cc5e13986f16e09df0
and
https://github.com/systemd/systemd-stable/commit/cc6271f17d3c5c200e8a391f10d0afb71403fc28
(as a de
Fwiw, the patch debian/lp1934147/0001-core-add-a-new-unit-method-
catchup.patch is part of a fairly large patchset
(https://github.com/systemd/systemd/pull/9200/commits) so maybe
something is missing there?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is su
** Merge proposal linked:
https://code.launchpad.net/~slyon/ubuntu/+source/systemd/+git/systemd/+merge/411528
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails
Thank you for testing the revert!
Do we know if the same issue happens in Focal/UC20 as well? Or can we somehow
falsify that? The regressing patches had been introduced to
Hirsute/Focal/Bionic to fix LP: #1934147
--
You received this bug notification because you are a member of Ubuntu
Bugs, whi
Thank you Michael for verifying this! As per my earlier question during
our last-week's sync: do you think it would be possible to have this
running for instance as part of the coreXX-edge testing that Sergio is
doing? Since we'd like to make sure we don't reintroduce this issue at
some point by ac
Just for the record, I also did another run with:
https://storage.googleapis.com/snapd-spread-
tests/snaps/core18_20211102_amd64.snap and that also fails right away:
"""
$ git status
On branch tests-use-core18-from-gce
Your branch is up to date with 'sergiocazzolato/tests-use-core18-from-gce'.
$ g
Thanks Lukas for providing this revert of
f0831ed2a03fcef582660be1c3b1a9f3e267e656. Using
https://people.ubuntu.com/~slyon/uc18/core18_20211105_amd64.snap I can
no longer reproduce the isue.
To make it easier to reproduce what I did:
"""
$ git status
On branch tests-use-core18-from-gce
Your branch
** Changed in: systemd (Ubuntu Bionic)
Assignee: (unassigned) => Lukas Märdian (slyon)
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount unit
Sounds like we cannot have an easy fix by simply applying
04eb582acc203eab0bc5c2cc5e13986f16e09df0 (and its dependencies).
So let's first try to pinpoint if f0831ed2a03fcef582660be1c3b1a9f3e267e656
really introduced the regression. I've built a new core18 snap with my previous
work and the inclu
** Also affects: systemd (Ubuntu Bionic)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ub
We created a test PR that uses this core18 snap build in
https://github.com/snapcore/snapd/pull/11015 - we are still running
tests but it seems the error is still there :-(
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.l
** Tags removed: rls-jj-incoming
** Also affects: systemd (Ubuntu Jammy)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fa
** Tags added: fr-1832
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubuntu Core 18
To manage notifications about this bug go to:
ht
Sure, I put that custom core18 snap into
https://people.ubuntu.com/~slyon/uc18/ as well.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units i
@slyon could you make available the core18 snap you built with this
systemd ?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubuntu Co
Not having a reliable reproducer is always a tricky situation to be
in...
Indeed that other commit might be related. I've backported the following
patches and created a custom systemd build in
https://launchpad.net/~slyon/+archive/ubuntu/lp1949089
https://github.com/systemd/systemd-stable/commit/
** Tags added: rls-jj-incoming
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1949089
Title:
systemd randomly fails to activate mount units in Ubuntu Core 18
To manage notifications about this bug g
42 matches
Mail list logo
