Since tests are launched inside namespace, we don't have SUID ping
available anymore so we explicitely allow all users to do ping.

Please note that this works for both IPv4 and IPv6 despite the name.

See nexedi/slapos!1107 (comment 148968)
for more information.

While bootstraping a host, the network can be in unknown state, and a
lot of tasks depend on the network, thus make sure that network is
available.

By extracting the hostname task to vm-bootstrap-hostname it's possible
to setup it as a dependency role for vm-bootstrap, before package role,
which depends on the network setup.

For KVM version < 1.0.220, the script /etc/opt/netconfig.sh is doing:

1) add address on IFACE
2) add routes going through IFACE
3) set IFACE up

But we can't add routes to an interface which is down. So call the
script twice so that the routes are really added during the second run.

The fix for KVM can be found at nexedi/slapos@07439e2e

slapos.cfg example is maintained in slapos.core and available at
/opt/slapos/eggs/slapos.core-*.egg/slapos/slapos.cfg.example

Also make vm-bootstrap role depends on python3

It was always done like this but has been removed in a previous cleanup (c7183d54). No need anymore for the upload.yml playbook.

In old Debian 10, after the release of Debian 11, apt update is failing
like this if we don't add the option "--allow-releaseinfo-change":

root@debian:~# apt update
Get:1 http://security.debian.org/debian-security buster/updates InRelease [65.4 kB]
Get:2 http://http.us.debian.org/debian buster InRelease [122 kB]
E: Repository 'http://security.debian.org/debian-security buster/updates InRelease' changed its 'Suite' value from 'stable' to 'oldstable'
N: This must be accepted explicitly before updates for this repository can be applied. See apt-secure(8) manpage for details.

See merge request nexedi/slapos.package!154

new one already added in dc173ee3

See merge request !155

See merge request nexedi/slapos.package!153

Same as nexedi/re6stnet@f1e95660

See merge request nexedi/slapos.package!151

On slower hosts it takes a bit more time to boot the ISO some times, which
results with not receiving the ESC key to disable the graphic screen, and so
not working installation.

The general idea is to provide a machine connected to re6st, registered as
computer in SlapOS Master and with thorwable user to work on it.

Starting from default installation some uneeded packages are removed, and
few expected ones are added in order to have common set of packages and use
scenarios across all installations.

It's not possible on Ubuntu/Focal

The goal is to use the newest re6st-node package that correctly enables
the re6stnet service. re6st-node package has been fixed with
re6stnet@5bdb32e6
and 6972e11f