Add mgmofs.protowfhostport mgmofs.protowfendpoint to MGM config before the MGM actually start and fix protowfendpoint: it must begin with '/'.

nothing...

Using /EOSOK file to check that eos instance has been launched successfully and not kubectl logs as there is no log with systemd.

services removed in the previous one.
The solution for now is to disable selinux on the host running the
containers.

systemctl is in a bad state... removing some services according to https://hub.docker.com/r/centos/systemd/~/dockerfile/

removing cta-cli as it has many dependencies that are breaking systemd start (in network.service startup)

Indeed if the derived containers are booted with selinux already installed systemd does not work because of different selinux policies on the same file in the container and the host system:
```
[root@ctafrontend ~]# systemctl
Failed to get D-Bus connection: Operation not permitted
```

Anyway even if selinux is installed later it is not enabled in the running container...

-/usr/sbin/initnone-
As I do not want a complex sed query to check for systemd or not, let's
solve it in init_pod.sh and express it in CI_CONTEXT variable.

/shared/rc.local in case I loose my devbox.

out if we are using systemd or not. We just need to check
/proc/1/cmdline to get the command that started the container.
Now to chose between systemd and no systemd we just need to set
`command` key in the pod yaml definition for nosystemd or remove the
line to use systemd.
Much more straightforward...

If CI passes...

[skip CI] New versions of kubernetes always pull the latest image, not matter what the PullPolicy is, the solution is to tag this image with a predefined tag. There could be contention here if we are launching 2 buildtree instances simultaneously but anyway, there is already a problem with the tag latest of this image...

-/usr/sbin/initnone-
As I do not want a complex sed query to check for systemd or not, let's
solve it in init_pod.sh and express it in CI_CONTEXT variable.

/shared/rc.local in case I loose my devbox.

out if we are using systemd or not. We just need to check
/proc/1/cmdline to get the command that started the container.
Now to chose between systemd and no systemd we just need to set
`command` key in the pod yaml definition for nosystemd or remove the
line to use systemd.
Much more straightforward...

If CI passes...

This reverts commit 421c6d14.

This version of EOS appears to work on my local dev box but not
in the CI.