[#165] Liveness probes experiment [DONT merge]

[SilviaAmAm]

Part of #165

To discuss with @sergei-maertens @alextreme @Tiho0135 :

• How do we feel about the maykinhelmlib?
• How do we feel about having the liveness probe script as a configmap like this? Advantage: we don't have to update all the projects, disadvantage: we need to update the ansible deploy

[alextreme]

Having a shared helmlib makes sense, there is a lot of duplication in the charts so I'm in favour of (step by step) converging the templates, although it may make how the charts work less obvious

I don't have an opinion about the liveness probe script in a configmap, I don't know what the best practices are

[sergei-maertens]

I'm out of my depth to give proper opinions on this, so I leave it for the devops team to come to a consensus.

The only thing catching my eye is that maykinhelmlib isn't particularly inspiring for a lib name. I think it's quite obvious that it's for helm, so repeating that name looks odd to me. Maybe something like maykin-utils or maykin-django-helpers is a better name?

[Tiho0135]

I fully support the helm library chart, very good suggestion by Silvia.
Using a ConfigMap for the liveness probe script is a normal way to be done. ConfigMaps can be treated as volumes.
I think the advantage is significant since it is implementing the DRY(do not repeat yourself) principle.