[systemd-devel] UID/GID broken on container

[arnaud gaboury]

I am refering here to issue #4078[0] and see if I missed any progress. I am
still with a broken UID on my container on some directories. Example:

>From inside the container:

---------------------------------------
$ ls -la /proc
....
-rw-r--r--   1 nobody          nobody             0 Dec 12 11:47 timer_stats
-r--r--r--   1 nobody          nobody             0 Dec 12 11:47 uptime
-r--r--r--   1 nobody          nobody             0 Dec 12 11:47 version
-r--------   1 nobody          nobody             0 Dec 12 11:47 vmallocinfo
-r--r--r--   1 nobody          nobody             0 Dec 12 11:47 vmstat
-r--r--r--   1 nobody          nobody             0 Dec 12 11:47 zoneinfo
-------------------------------------------

I usually use fuidshift command on the host to fix the wrong UID/GID, and
it globally works. But I can't fix  the /proc dir in the container as the
host do not see it.

Can anyone help me in fixing this issue on /proc? I need to upgrade the
container (Fedora) but can't because of the nobody:nobody owner.

In general, user name space is configured in my kernel, and systemd-nspawn@
is the original one, with the -U option. Am I right to think this is the
correct setup ?
Any news on this UID/GID issue?

TY for help


[0]https://github.com/systemd/systemd/issues/4078
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/systemd-devel/attachments/20161212/8c57250c/attachment.html>