Join GitHub today
GitHub is home to over 20 million developers working together to host and review code, manage projects, and build software together.
lxc-start: segfault #1280
Comments
|
does this also happen with privileged containers? |
ne0zer0
commented
Nov 6, 2016
|
Here what I get with privileged container:
|
ne0zer0
commented
Nov 6, 2016
|
Find also this in journalctl:
|
|
can you please try removing cgmanager, rebooting and re-trying? |
ne0zer0
commented
Nov 6, 2016
|
Done, here the output:
|
|
This strongly reminds me of: #1258. In any case, the segfault you're seeing should be fixed in current git master as soon as @hallyn finds the time to look at #1262.
makes this especially likely. In any case,
Please report back if the issue still persists. If it does, please show the output of:
I might need to request further information later on. :) |
ne0zer0
commented
Nov 6, 2016
|
The issue still persists:
zero@debian:~$ cat /tmp/lxc_test.log |
|
Apart from the segfault that is fixed by a recent pr that yet has to be merged I mentioned you are not placed in a writeable cgroup for crucial subsystems such as |
ne0zer0
commented
Nov 6, 2016
|
Already done:
but the issue still persists :s |
|
Did you install |
ne0zer0
commented
Nov 6, 2016
|
Of course: |
|
So
|
ne0zer0
commented
Nov 6, 2016
|
/etc/pam.d/common-session
/etc/pam.d/common-session-noninteractive
|
|
Looking at the output of your
means that the |
brauner
closed this
Nov 6, 2016
ne0zer0
commented
Nov 7, 2016
|
Is there any roadmap for cgroup v2? Because since Linux kernel 4.5 and Systemd 232, cgroup v2 are set by default. I thought that was since libvirt0 update, that lxc stopped to work, but the problem came with the update of systemd 231-10>232-1. |
|
It seems the current workaround is to boot with |
ne0zer0
commented
Nov 7, 2016
|
Thanks, I will try this. And, can you give me a link? |
|
So far the only documentation on that is the PR which implemented it: systemd/systemd#3965 |
ne0zer0
commented
Nov 7, 2016
|
Super! |
phedders
commented
Nov 8, 2016
|
systemd 232 broke lxc, docker and rkt for me. Not a pleasant surprise. |
|
Sounds like someone jump the gun here, distros very much shouldn't be switching to cgroupv2 until it's at parity with cgroupv1, and that won't be true until it's got a cpu controller... Hopefully distros will be fixing this quickly by just forcing systemd in cgroupv1 mode until that happens... |
|
This systemd/systemd#3965 really is a bit premature. |
|
At least it's limited to the systemd hierarchy, but that's still going to be a big problem for anyone who cares about running containers that use systemd as their init systems... |
evverx
referenced this issue
in systemd/systemd
Nov 8, 2016
Merged
Using the unified hierarchy for /sys/fs/cgroup/systemd when legacy hierarchies are being used #3965
brauner
reopened this
Nov 9, 2016
added a commit
to martinpitt/systemd
that referenced
this issue
Nov 9, 2016
martinpitt
referenced this issue
in systemd/systemd
Nov 9, 2016
Merged
core: don't use the unified hierarchy for the systemd cgroup yet #4628
added a commit
to systemd/systemd
that referenced
this issue
Nov 10, 2016
|
systemd/systemd#4628 reverted this for now, but you can still boot with |
phedders
commented
Nov 10, 2016
|
Many thanks. On 10 Nov 2016, 07:11, at 07:11, Martin Pitt notifications@github.com wrote:
|
added a commit
to fbuihuu/systemd-opensuse-next
that referenced
this issue
Nov 21, 2016
|
Closing this issue and opening #1316 to track cgroupfs v2 support. |
ne0zer0 commentedNov 6, 2016
•
Edited 1 time
-
ne0zer0
Nov 6, 2016
Hi,
Seems since the latest libvirt0 and libvirt-* packages update (2.3.0-3 > 2.4.0-1), I can not start anymore lxc.
zero@debian:~$ lxc-start --name ubuntu
Here debug